Renewable and Sustainable Energy Reviews 135 (2021) 110223

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: http://www.elsevier.com/locate/rser

A review of industry 4.0 revolution potential in a sustainable and

renewable palm oil industry: HAZOP approach
Chun Hsion Lim a, *, Steven Lim a, Bing Shen How b, Wendy Pei Qin Ng c, Sue Lin Ngan d,
Wei Dong Leong e, Hon Loong Lam e
a
Universiti Tunku Abdul Rahman, Department of Chemical Engineering, Jalan Sungai Long, Bandar Sungai Long, Cheras, 43000, Kajang, Selangor, Malaysia
b
Swinburne University of Technology Sarawak, Research Centre for Sustainable Technologies, Faculty of Engineering, Computing and Science, Jalan Simpang Tiga,
93350, Kuching, Sarawak, Malaysia
c
Curtin University Malaysia, Department of Chemical Engineering, CDT 250, 98009, Miri, Sarawak, Malaysia
d
UKM-Graduate School of Business, Universiti Kebangsaan Malaysia 43600 UKM, Bangi, Selangor, Malaysia
e
The University of Nottingham Malaysia Campus, Department of Chemical Engineering, 43500, Jalan Broga, Semenyih, Selangor, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: Palm oil is a renewable resource that has the potential to replace fossil fuel and petrochemical for a better
Palm oil sustainable system. However, there is room for improvement in the current operation of the palm oil industry to
Industry 4.0 achieve better sustainability development. The industrial revolution toward automation and artificial intelli­
HAZOP literature review
gence (AI) is the new trend known as the fourth industrial revolution (Industry 4.0). Unfortunately, the palm oil
Certified sustainable palm oil
Smart palm oil process
industry has been moving slowly in this revolution. This paper aims to conduct a detailed review of the current
Palm oil 4.0 state of the palm oil industry development toward Industry 4.0. A novel Hazard and Operability Analysis
(HAZOP) approach is adopted i) to ensure a detailed evaluation of the existing problems, and ii) to identify
potential implementation of Industry 4.0 technologies in the palm oil industry. HAZOP is a common approach
used in chemical engineering to systematically evaluate process safety and identify the possible improvement of
the existing system. The same concept is applied in this paper to investigate the possible adaptation of Industry
4.0 technologies to improve the palm oil industry. Existing Industry 4.0 technologies and features were evaluated
to identify feasible adaptation in the industry. The HAZOP review proposed 23 recommendations to improve the
palm oil industry with Industry 4.0 technologies to achieve a higher standard in sustainable production. A total of
13 specific Industry 4.0 features were identified as the potential development gaps for palm oil industry
stakeholders, which included the adaptation of Internet-of-Things sensors, cloud computing, blockchain, and
smart imaging processing technologies.

1. Introduction
The rapid development of industrialisation to improve the quality of
life, especially to meet the increasing demand of energy consumption
has significantly increased pollution. Recent developments focus on
reducing the dependence on fossil fuel by replacing it with other
environmental-friendly and renewable resources such as solar, wind and
geothermal. Significant research has been conducted to improve
renewable energy production to achieve the sustainable goal initiated by
the European Union to reduce 20% of the carbon dioxide and energy
consumption by 2020 [1]. Being one of the primary sources for edible
oil, chemical product and energy, palm oil industry has been growing
rapidly due to the increasing demand. Global palm oil production in­
creases tremendously in the past decades, almost doubling every 10
years [2]. Reported global industry consumption of palm oil and palm
kernel oil was about 19.72 million tons in 2018, while the highest
producing country, Indonesia, alone aims to produce 40 million tons by
2020 [3,4]. A study shows that palm oil is an energy crop which has the
highest oil production yield as compared to other oil crops [5]. This
enables optimum usage of limited plantation site (or land) to generate
the highest amount of oil for food, chemical and energy applications.
Apart from that, wastes generated from palm oil industry such as palm
biomass and palm oil mill effluent have shown to be a promising

* Corresponding author.
E-mail addresses: chlim@utar.edu.my (C.H. Lim), stevenlim@utar.edu.my (S. Lim), bshow@swinburne.edu.my (B.S. How), wendyngpq@curtin.edu.my
(W.P.Q. Ng), suelin.ngan@ukm.edu.my (S.L. Ngan), leongweidong@gmail.com (W.D. Leong), HonLoong.Lam@nottingham.edu.my (H.L. Lam).

https://doi.org/10.1016/j.rser.2020.110223
Received 8 February 2020; Received in revised form 6 August 2020; Accepted 6 August 2020
Available online 24 August 2020
1364-0321/© 2020 Elsevier Ltd. All rights reserved.
C.H. Lim et al.
List of abbreviations
AI Artificial intelligence
CPO Crude palm oil
CSPO Certified Sustainable Palm Oil
GPS Global Positioning System
HAZOP Hazard and Operability analysis
IoT Internet-of-Things
MSPO Malaysia Sustainable Palm Oil
RGB Red, green, blue
RSPO Roundtable on Sustainable Palm Oil
SMEs Small and medium enterprises
alternative for renewable energy generation. For example, palm biomass
can be used in pyrolysis, gasification, combustion, and liquefaction
process to generate energy [6].
Despite being a promising renewable resource, a study shows that
the palm oil industry has caused adverse environmental impacts, such as
deforestation, habitat loss, pollution (water and air) and forest fire [2].
With the constant international pressure and heightening demand for
sustainable palm oil production, the Roundtable on Sustainable Palm Oil
(RSPO) standard was established to promote the production of Certified
Sustainable Palm Oil (CSPO) for environment protection [7]. Being the
major palm oil producers, Malaysia and Indonesia have initiated
Malaysia Sustainable Palm Oil (MSPO) and Indonesia Sustainable Palm
Oil (ISPO) schemes respectively. However, regardless of numerous ef­
forts and awareness campaigns to promote MSPO and ISPO, the adap­
tation of CSPO production was slow, especially among small and
medium enterprises (SMEs). Researchers highlighted several reasons for
CSPO implementation failure, including premium for the CSPO is too
low compared to the implementation cost, as well as the complex system
to trace and differentiate CSPO and non-certified palm oil [8]. Another
study suggested that certified palm oil production is not as sustainable as
it claimed as the certification process might cause more tree losses [9].
Numerous efforts have been contributed by researchers to improve
the RSPO system from different aspects. For example, RSPO evaluation
method was discussed and enhanced with a new and improved quanti­
tative assessment tool based on air, water, and soil criteria [10]. Alter­
native standard schemes for sustainable palm oil were evaluated to
suggest that the future development of the industry should be focusing
on increasing participation of smallholders in CSPO certification [11].
Apart from the efforts to achieve certified production, other researchers
have contributed to improve sustainability practice in oil palm tech­
nologies. Multi-objective optimisation model that incorporated stake­
holders’ and experts’ value judgement was proposed to improve the
planning of sustainable oil palm value chains [12]. The carbon footprint
sustainability index in palm oil mills was evaluated to measure the
carbon emissions at each stage of the process [13]. The multiple oil crops
supply chain system was considered to optimise the overall sustain­
ability profile of the oil crop processes [14]. Noble metallic catalysts
were used in the production of renewable diesel from palm oil as an
alternative energy source to replace fossil diesel [15]. Researches have
shown that a carbon emission reduction of up to 73% can be achieved by
substituting the conventional fossil diesel with the palm-derived bio­
diesel [16,17]. Due to the high generation yield, oil palm is classified as
the more effective crop in terms of land utilisation, productivity and
marketability, which may help to reduce dependency on fossil fuel [18].
Despite that, the European Parliament has voted to ban the usage of
palm oil in biodiesel production in the effort to stop deforestation in
tropical countries [19]. This further emphasised the importance of
achieving and proving that there is sustainable palm oil production for
continuous usage and exportation of palm oil-based product.
There are still many challenges in the implementation of the RSPO
2
Renewable and Sustainable Energy Reviews 135 (2021) 110223
standard in the palm oil industry. Issues such as lack of information
transparency, traceability, the low and unfair premium price for CSPO
are some of the major factors that discourage the initiation of CSPO
production [20]. Currently, the world is moving towards the Industry
4.0, which focuses on the digital transformation of the industry
including the development of automation, smart system, digitalisation,
big data analysis, Internet-of-Things (IoT) and artificial intelligence (AI).
Industry 4.0 is first introduced at Hanover Fair by Germany as the
concept of technology-driven manufacturing process integrated with
information and communication technologies to strengthen its national
competitive position in manufacturing [21]. Various researches were
discussed in a study to incorporate Industry 4.0 to enhance sustainability
performance [22]. Implementation of Industry 4.0 concept and tech­
nology may be a potential solution to improve the effectiveness of CSPO
execution. However, the implementation of such ideas requires a robust
and smart system to handle the massive amount of data and variables.
There is still a huge gap for the palm oil industry to fully adapt to In­
dustry 4.0, primarily due to the absence of proper guidelines and
reference cases for stakeholders at the current stage.
The focus of this paper is to review the palm oil industry from up­
stream processes (plantation and mill) to downstream markets (food,
energy and chemical products) to evaluate the potential implementation
of Industry 4.0 technologies in each sector. As Industry 4.0 is relatively
new to the palm oil industry, the proposed review process considered
any form of Industry 4.0 technologies and development (in palm oil or
other industries) that is readily available. The main objective is to
identify possible solutions to improve existing palm oil industry pro­
cesses with Industry 4.0 technologies. To the best of authors’ knowledge,
no review has been conducted to relate Industry 4.0 technologies and
features with the palm oil industry for food, energy and chemical pro­
duction. Due to the newness of the scope with lack of real industry
application, a brain-storming exercise is adapted in this research with
literature support based on Industry 4.0 technologies from other in­
dustries and input from the palm oil industry experts. In order to ensure
a systematic and all-rounded review, Hazard and Operability Analysis
(HAZOP) methodology is adopted. HAZOP is a systematic brain-
storming approach used in chemical engineering i) to identify poten­
tial process hazard, ii) to assess the ultimate consequence of the cause,
iii) to evaluate the adequacy of existing safeguards and mitigations, and
iv) to propose recommendations for improvement if required [23]. This
approach has shown great potential to identify research gaps and po­
tential implementation in biomass processes [24]. In short, a detailed
review of the current palm oil industry is conducted in this work with
HAZOP literature review approach by i) identifying the credible process
deviations in the palm oil industry, ii) assessing the respective conse­
quences, iii) evaluating current practices to prevent the deviations or to
mitigate the consequences, and finally iv) evaluating any possible In­
dustry 4.0 technologies adaptation that can help to rectify the problems
in the palm oil industry.
2. Methodology
In this paper, the HAZOP methodology is adopted to identify the
potential implementation of Industry 4.0 technologies into the palm oil
industry. The HAZOP approach is used in various industries to evaluate
process safety and identify potential improvement. The oil and gas in­
dustry is actively applying this method to evaluate and improve process
safety and control based on piping and instrumental diagrams (P&IDs).
Typically, a third party is engaged in the review process to ensure an
unbiased evaluation. A study was conducted to compare the industry
HAZOP approach and the HAZOP literature review approach, and has
successfully incorporated the HAZOP approach into biomass supply
chain optimisation model literature review [24]. The same methodology
is adopted in this paper to review the possibility of integrating Industry
4.0 technologies into palm-based food, energy and chemical production
processes. Fig. 1 shows the proposed procedure for the modified HAZOP
C.H. Lim et al.
literature review approach designed for this paper.
As per the first step in the procedure, Fig. 2 represents a simplified
block flow diagram of the palm oil supply chain used for this study. The
overall process was classified into 3 nodes to enable a focused review
process. The deviations and guidewords that used to trigger imagination
and brainstorming of potential process deviation/hazard in this study
are summarised in Table 1. Several assumptions and limitations have
been listed prior to the review process, which include i) only single
failure/deviation is considered at any time, ii) any form of Industry 4.0
technologies (implemented in existing palm oil and other industries, and
developing technologies) are considered in this study, and iii) any form
of input (literature, interview, survey, etc.) from industry stakeholders
are considered as credible source and information. The HAZOP review
was conducted to the best knowledge of authors based on available in­
formation from literature and input from palm oil industry stakeholders.
Deviations and guidewords are used to trigger a structured imagination
of possible process upsets or problems. Credible causes and their
respective ultimate consequences are also discussed and summarised in
worksheets (see Appendix). Upon assessing the problems and conse­
quences in the industry, the team also conducted research based on
literature and industry input to identify existing practices or methods
that could mitigate/safeguard the problems. The identified problems
were related to existing Industry 4.0 technologies and features (that
have yet to be implemented in the palm oil industry) to access the po­
tential adaptation which may help to minimise/prevent the problems.
The verification of existing technologies was mainly focused on the
publications (within 5 years) from the Elsevier database, while some
older publications were included if the content is highly relevant to this
work [25]. The screening of the literature was conducted based on the
keywords identified from the HAZOP recommendations, including palm
oil, IoT, weather forecast, soil detection, AI, deep learning, image
Fig. 1. HAZOP literature review procedure.
3
Renewable and Sustainable Energy Reviews 135 (2021) 110223
processing, cloud computing, oil palm ripeness, blockchain, cloud
computing, smart contract, cybersecurity, big data analysis, and opti­
misation. Recommendations were suggested to highlight the possible
integration of Industry 4.0 technologies into the palm oil industry as
research and development gaps. These recommendations can be adop­
ted by stakeholders as a guideline to study the possible Industry 4.0
implementation and investment for their respective sectors.
3. Results and discussion
A comprehensive HAZOP review was conducted based on the scope
presented in Fig. 2. The details of the discussions were recorded and
documented in Table A.1, A.2 and A.3 in the appendices. Table 2 sum­
maries all the recommendations proposed from the HAZOP review.
These recommendations are the compiled information for potential
improvement and implementation of Industry 4.0 technologies into the
palm oil industry based on the current industry problems and maturity
of Industry 4.0 technologies. A total of 14 recommendations were
identified from the upstream processes (Node 1), 8 new recommenda­
tions proposed for downstream processes (Node 2) and 1 new recom­
mendation was introduced from market/retailers’ point of view (Node
3). Analysis of the outcomes suggested that some recommendations
were interrelated and addressing a similar issue in the palm oil industry,
which can be rectified or improved with Industry 4.0 technologies. A
total of 8 classes of Industry 4.0 technologies were identified that have
the potential to be implemented in the palm oil industry. The following
sections discussed the details of each technology and its respective
function in the palm oil industry.
C.H. Lim et al.
Fig. 2. Node classification for palm oil processes.
3.1. IoT devices-based implementation
Upon analysis and investigation, Recommendations 1 and 2 (sum­
marised in Table 2) proposed from the HAZOP review can be summar­
ised as potential gaps to utilise IoT devices to improve oil palm
plantation management. Recommendation 1 (Incorporation of localised
IoT devices to improve the accuracy of weather prediction) suggested
that IoT devices may help to minimise the impact on production yield
and activities due to uncertain weather conditions. Based on the oil palm
plantation expert’s comments, harvesting of fresh palm fruit is highly
dependent on weather conditions. Currently, daily harvesting activity is
planned according to the regional or national weather forecast, which
does not reflect the actual condition at the plantation site. Thus, local­
ised devices to detect plantation site parameters for an improved
weather prediction system may help to increase the efficiency of plan­
tation management. A study has shown poor labour planning due to
inaccurate weather forecast may cause a delay in harvesting, thus
4
Renewable and Sustainable Energy Reviews 135 (2021) 110223
subsequently reduces the oil extraction and profit [26]. Various re­
searches were conducted to enhance the local weather prediction model.
For instance, an adaptive weather forecasting device based on local
temperature, wind speed and global radiation was proposed for agri­
culture application [27]. The proposed model has shown better accuracy
in providing short-term forecasts using local devices, including predic­
tion of seasonal changes based on weather conditions in previous years.
Convolutional neural network architecture which capable of conducting
accurate weather forecasts based on the collected regional temperature
and wind speed to improve the prediction performance [28]. Literature
has shown that IoT devices can be implemented into upstream of the
palm oil industry to have a better local weather forecast system for more
stable harvesting and production planning.
Similarly, Recommendation 2 (Incorporation of localised IoT devices
to detect soil properties for moisture and nutrient) addressed the need
for local devices to detect and control soil properties, such as nutrients
and moisture for optimum plantation conditions. Soil nutrients may be
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223

Table 1 balanced nutrient regime for palm tree has shown better yield, espe­
Deviations and Guidewords for HAZOP review. cially among small stakeholders [30]. Many efforts have been conducted
Deviations Guidewords to evaluate the impacts of fertiliser usage for higher oil yield and better
management practices [26,31–33]. Besides, annual soil testing and
Feed Fluctuation of quantity
Inconsistency of quality fertiliser planning are required to maintain good soil condition. Based on
Storage the oil palm plantation practice in Malaysia, soil analysis and fertiliser
Various suppliers application are typically conducted between January and February to
Processing efficiency Monitoring of process avoid monsoon season (minimise nutrient loss); while a corrective fer­
Process intervention/control
Maintenance
tiliser application is conducted between July to September. This process
Product Fluctuation of quantity may take up to a few weeks with the involvement of intensive labour
Inconsistency of quality effort to collect a huge number of soil samples across the plantation sites.
Storage A study utilised principal component analysis to demonstrate that geo­
Product distribution
metric enzyme activity, available phosphorus and moisture content are
Cost Fluctuation of feed cost
Fluctuation of product price the key factors to determine soil quality in oil palm plantation [34]. This
Processing cost study concluded that soil testing analysis could be reduced to only three
Accuracy of information Traceability of resources parameters. However, despite the potential to develop a smart IoT sys­
Information security tem to monitor soil conditions, the advantages of this feature may be
insignificant due to the relatively low demand and frequency of soil
testing and fertiliser activity requirement throughout the production
Table 2 years. A detailed cost and benefit analysis should be conducted prior to
List of recommendations. the implementation to compare the functionality of the automation soil
Recommendations detection system with the manual soil detection system, especially for
the application in SMEs.
Node 1:
1. Incorporation of localised IoT devices to improve the accuracy of weather A smart plantation system is a more feasible implementation with
prediction. consideration of multiple factors, include water and soil management to
2. Incorporation of localised IoT devices to detect soil properties for moisture and reduce the dependency of labour which improves efficiencyand mini­
nutrient.
mises cost. The concept of smart farming is not new to the world as
3. AI imaging processing technology to determine tree health status locally.
4. Health care devices such as smartwatches to monitor health status and location of
various researches have been conducted. A comprehensive review and
worker. comparison of multiple IoT applications in smart farming have been
5. AI image processing technology to determine the ripeness of palm fruit locally. discussed elsewhere [35]. Unfortunately, to date, there is no
6. Combined IoT device to detect ripped fruit and cloud computing to optimise one-size-fits-all protocol that can perform well in all aspects including
harvesting plan/route to save time.
latency, energy and bandwidth requirements, throughput, reliability
7. Incorporate IoT and big data analysis on machinery, detectors, controllers to
determine optimum repair schedule prior to breakdown. and security. In the view of improving sustainable practice, fuzzy logic
8. IoT and big data analysis to improve forecasting techniques for fresh palm fruit control was incorporated into IoT devices to demonstrate a smart water
generation. irrigation system [36]. The work showed that implementing IoT in the
9. Blockchain technology to enhance the traceability of products across the supply system can reduce water usage and labour cost. A similar study in an
chain.
10. Cloud computing to optimise supply chain network from multiple sources; while
urban application of lawn management also demonstrated the advan­
all the required quantity, properties, and sustainability profile are achieved. tages of a smart irrigation system in achieving a more sustainable and
11. Smart contract to determine the premium price for good quality products. environmental-friendly implementation [37]. Researchers concluded
12. IoT devices and GPS system to detect live loading weight and truck location that the plant water deficit index threshold could be adopted into IoT
during distribution.
devices for smart irrigation scheduling and plant transpiration [38].
13. Provide tools for the digitalisation of printed documents.
14. Improve cybersecurity, such as blockchain for decentralised network. Apart from the smart irrigation system, the application of IoT was also
Node 2: adopted for real-time monitoring and control systems in a greenhouse
15. Improvement of CPO production forecast for more efficient supply chain and [39]. Despite the advantages of IoT features, large scale implementation
storage management. of IoT is still considered a challenging task, especially in oil palm
16. Incorporate smart contract to increase the selling price of fresh palm fruit for
SMEs with consistent supply.
plantation management due to lack of internet coverage in many plan­
17. Incorporation of IoT devices in the process to enable remote monitoring and tation sites and the network speed limitation. Apart from that, IoT de­
access for process control. vices and sensors should be placed strategically on site for optimum
18. Incorporation of big data analysis, cloud computing and IoT to enable detection coverage and cost. Nonetheless, the advancement of internet
communication between multiple control loop/system for mitigation action upon
and computing services such as 5G network will drive the IoT ecosystem
any control loop failure.
19. Incorporate blockchain and IoT technology to ensure each product is distributed in the business model of a smart enterprise [40]. To note, investment
within the maximum allowable storage time. cost of IoT devices should be emphasized in the development phase,
20. Incorporation of live weather forecast, traffic information and GPS to optimise especially with the consideration of relatively low investment budget
distribution network. from SMEs.
21. Big data analysis to forecast market demand for each downstream product to plan
for production strategy in advance.
22. Big data analysis to improve CPO generation forecast including weather and 3.2. Smart forecasting-based implementation
process conditions to enable more accurate prediction for pricing.
Node 3: Several requirements for better forecasting tools were identified
23. IoT and image processing devices to keep track of daily stock count.
from the HAZOP review in the aspects of weather, maintenance, pro­
duction rate and quality, market demand and pricing as shown in Rec­
lost through erosion of soil sediments, leaching, runoff (with rainwater), ommendations 1 (Incorporation of localised IoT devices to improve the
and fresh fruit bunch harvesting activity. Typically, empty fruit bunches accuracy of weather prediction), 7 (Incorporate IoT and big data anal­
are used as mulches to cover the soil around palm trees to preserve ysis on machinery, detectors, controllers to determine optimum repair
nutrients. Research has reported that the EFB mulch could help to in­ schedule prior to breakdown), 8 (IoT and big data analysis to improve
crease pH, Nitrogen, Phosphorus, and Potassium content [29]. A forecasting techniques for fresh palm fruit generation), 15 (Improve­
ment of CPO production forecast for more efficient supply chain and

5
C.H. Lim et al.
storage management), 21 (Big data analysis to forecast market demand
for each downstream product to plan for production strategy in
advance), and 22 (Big data analysis to improve CPO generation forecast
including weather and process conditions to enable more accurate pre­
diction for pricing) in Table 2. As discussed in the previous section, the
dynamic nature of palm oil production is directly affecting the pricing
and downstream applications, especially the competitive price com­
parison between biodiesel and petrodiesel. Therefore, the accuracy of
forecasting technology to predict the dynamic production and price of
palm oil is crucial for the industry to have more precise management and
planning. A series of Industry 4.0-based forecasting technologies are
discussed in this section, e.g., the incorporation of big data analysis and
deep machine learning. Study shows that a convolutional neural
network model can improve forecasting accuracy due to its ability to
capture and utilise multi-scale data characteristics [41]. Various ma­
chine learning algorithms based on precipitation, maximum and mini­
mum temperature data were applied to model climate changes in
Pakistan including artificial neural network, K-nearest neighbour, sup­
port vector machine and relevance vector machine [42]. An accurate
weather forecast provides a better basis of water and fertiliser man­
agement, and hence improves the production yield of palm oil. Re­
searchers adapted Bayesian networks to predict oil palm plantation
yield, which shows comparable results to the artificial neural network
approach [43].
Forecasting technology can also be applied in the process control and
management in the palm oil industry, such as palm oil mill and biodiesel
refinery. The need for a preventive measure in the processing was
highlighted in Recommendation 18 (Incorporation of big data analysis,
cloud computing and IoT to enable communication between multiple
control loop/system for mitigation action upon any control loop failure).
A comprehensive review was conducted on various AI applications in
the management of process maintenance [44]. The study also showed
that the genetic algorithm and neural network are the best approaches
for preventive maintenance planning and scheduling purpose. Another
work was conducted to predict the requirement of maintenance on
machine tools using AI techniques based on the machine condition data
to reduce unplanned downtime or failure [45]. This technology can be
implemented into the palm oil processes, equipment maintenance
detection and scheduling system to minimise equipment breakage. For
example, detection of mechanical failure in the oil extraction process
based on the movement of the pressing machine and prediction of the
maintenance schedule for the process control valve in refinery based on
the response time degradation. In terms of supply chain and logistics
management, researchers have improved the Baltic Dry Index fore­
casting accuracy with AI method to minimise risk due to changes in
shipping freights [46].
Implementation of Industry 4.0 technologies such as big data anal­
ysis, neural network and AI have shown the potential to improve the
accuracy in forecasting technology. Various researches have shown that
adaptation of such technology is possible in plantation, process, and
logistics. Nonetheless, for such technology to be integrated into the palm
oil industry, significant improvement of the existing equipment and
facilities are required. For instance, IoT-enabled sensors have to be
installed in plantation sites and processing plants for big data collection
and input into the forecasting technology. This may impose challenges
for stakeholders, especially SMEs, in terms of affordability due to
various uncertainties in revenue and cost [47].
3.3. Smart image processing technology-based implementation
From Recommendations 3 (AI imaging processing technology to
determine tree health status locally), 5 (AI image processing technology
to determine the ripeness of palm fruit locally), 13 (Provide tools for the
digitalisation of printed documents), and 23 (IoT and image processing
devices to keep track of daily stock count) in Table 2, the HAZOP review
shows that the advancement of image processing tool may help to
6
Renewable and Sustainable Energy Reviews 135 (2021) 110223
improve the efficiency of processes in the palm oil industry. In the
plantation site, identification of palm tree health status and fruit ripe­
ness has been one of the major process bottlenecks which involve
intensive labour and time. Based on oil palm plantation expert, satellite
imaging technology can be used to identify unhealthy trees (based on
leaf colour) within a large plantation site. Nonetheless, local observation
is still required for confirmation, especially for the case where a single
(or a small number of) unhealthy tree is surrounded by many healthy
trees. Similar smart imaging technology can be implemented for palm
fruit ripeness detection. Variation of flowering and ripeness of each palm
fruit in the same brunch or tree is very common due to the non-
homogenous pollination nature of oil palm [48]. This requires farmers
to identify ripped fruits visually at the site for harvesting purpose. Such
requirement and practice demand tremendous labour force; and poses a
relatively low efficiency. One of the potential adaptations of imaging
processing in an oil palm plantation is to use localised imaging system
such as CCTV or mobile imaging devices (mobile robotic or drone) for
the oil palm ripeness detection. The system should consider the uneven
terrain and lighting at the plantation site. However, to the authors’ best
knowledge, such a system has yet to be implemented in the palm oil
industry.
Despite that, on-going research and development to integrate smart
image processing technologies into oil palm process had been conduct­
ed. A mobile robot was built to scan and identify palm tree trunk in
plantation sites using colour image processing and depth data processing
techniques [49]. Detection of palm fruit ripeness by the red, green and
blue (RGB) colour unit from the image captured was investigated [50].
Similarly, a low-cost oil palm ripeness detector was constructed using a
digital camera and laptop to detect RGB of palm fruit placed in a small
chamber [51]. On the other hand, the incorporation of smart imaging
detection and computational power is required to accurately detect and
predict ripeness of palm fruit for harvesting planning. Study shows that
palm fruit ripeness can be detected with genetic algorithm neural
network based on near-infrared spectral data [52]. An artificial neural
network model was developed to detect ripeness of fresh fruit bunch via
hue, saturation and intensity colour model. The result shows that the
accuracy of the artificial neural network model is higher as compared to
the linear regression model [53]. A deep learning model to classify tree
species can be modified to classify palm trees (with fruit and without
fruit) for harvesting purpose [54]. Further improvement of the smart
imaging processing technology can be achieved by integrating the
detection system with IoT and cloud computing as proposed in
Recommendation 6 (Combined IoT device to detect ripped fruit and
cloud computing to optimise harvesting plan/route to save time). Un­
healthy trees or palm fruit ripeness can be detected with IoT-enabled
detection robot or drone where the information can be processed in a
centralised cloud computing server for analysis. For instance, the smart
detection system can be used to propose an optimum harvesting route
for farmers to reduce the harvesting period or to recommend an opti­
mum schedule for the tree health check.
Apart from oil palm plantation site, the implementation of image
processing technology also can be applied in palm oil processing such as
biodiesel refinery. One of the applications includes detection of defects
on process equipment and product to improve the efficiency of main­
tenance activities and quality checking processes. Research has suc­
cessfully developed a deep learning model to detect piping defect via
CCTV footage [55]. Implementation of such a detection system may help
to prevent unpredicted piping failure, especially for hazardous process
piping and equipment. A comprehensive review of the application of
non-destructive sensors to detect horticultural products and the appli­
cation of machine learning was conducted as part of the detection tools
to improve the accuracy of the system [56]. Image processing technol­
ogy can also be applied in the retailer operation as suggested from
HAZOP Recommendation 23 (IoT and image processing devices to keep
track of daily stock count). For instance, unmanned retailer shop or
stock-check process using CCTV and sensor to detect changes of palm oil
C.H. Lim et al.
product on the shelve. This idea has been demonstrated in a study of
real-time recognition system to detect and track products in the cabinet
[57]. Complex smart tracking technology can further be improved by
incorporation of demand forecasting based on the product flow detec­
tion on the shelve to manage stock replenishment accurately. For
example, the production rate of palm biodiesel is planned according to
the demand of biodiesel to avoid over or under production to maintain a
healthy biodiesel price.
Development of image processing technology helps in the digital
transformation of the palm oil industry and its operations. Information
transfer and documentation such as specification sheet (or datasheet) of
palm oil product, checklist for maintenance, etc., are usually in hard­
copies, especially among SMEs. One of the reasons that many stake­
holders are reluctant to put extra effort into Industry 4.0 is due to the
tedious transition period, such as digitalisation of system that includes
converting historical data from hardcopy into softcopy. This creates an
additional barrier for the advancement of the palm oil industry into the
Industry 4.0 era. The issue was highlighted in Recommendations 13
(Provide tools for the digitalisation of printed documents) where a smart
system is needed to aid the transformation towards the paperless system.
This paper has identified several researches to improve the efficiency of
translating hardcopy documents into softcopy via image processing. For
example, a deep learning system was constructed to detect row of sen­
tences within a written document for better digital translation results
[58]. Similar work was conducted to recognise different languages to
cater to different nationalities of workers. For example, English hand­
written word image identification was conducted based on Arnold
transform approach to recognise and record written documents in soft­
copy, while similar work was undertaken to recognise written Chinese
characters [59,60]. Adaptation of such technology may help to promote
more initiatives from palm oil industry stakeholders to move towards
Industry 4.0. Nevertheless, the system should uphold a proper security
system to avoid leakage of sensitive information. This issue will be
addressed with cybersecurity in Section 3.8.
3.4. Cloud computing-based implementation
Previous sections had discussed various potential implementations of
Industry 4.0 technologies in the palm oil industry, including localised
IoT devices, forecasting and smart image processing. The operation of
such technologies requires a considerable amount of computing power
for big data analysis and AI processing. This creates a challenge among
palm oil industry stakeholders, especially SMEs, which do not have
advanced computing facilities in their conventional operation. Cloud
computing technology is one of the Industry 4.0 technologies which
enables multiple users to access a centralised powerful computing ma­
chine with dynamic and reconfigurable computing applications. The
applications include remote access for information sharing, editing and
storage, database management, and processing large datasets for com­
plex problems [61]. There are various types of cloud computing,
including private cloud, public cloud, community cloud and hybrid
cloud with their respective features and applications [62]. The main
advantage of cloud computing is that it allows users to send the raw data
remotely to the cloud server for complex data processing, then the
processed result is delivered back to the users. This enables users to
access powerful computing facility with a minimum cost of subscription
for the service to avoid huge investment cost into physical computing
facility. This feature could be a more favourable option to the palm oil
industry, especially SMEs in the plantation. A study has shown that
migration to the cloud server for data storage and computing helps the
organisation by reducing capital and operational cost, improve flexi­
bility, and increase revenue, especially among SMEs [63]. Nonetheless,
accessibility of cloud computing depends on internet access and con­
nectivity, which is a common facility readily available in the palm oil
industry, except in rural plantation sites.
Various applications of cloud computing have been implemented
7
Renewable and Sustainable Energy Reviews 135 (2021) 110223
across many industries and problems. A study was conducted by using a
free cloud computing platform and image processing technology from
Google Earth Engine to classify the oil palm plantation map into water,
built-up, forest, bare soil, oil palm, paddy and other vegetation [64].
Nonetheless, distinguishing oil palm plantation with other vegetation
from Google Earth image has proven to be a challenging work due to the
similarity in spectral image and structure [65]. Complex process opti­
misation in palm oil processing plants can be solved with cloud
computing to reduce the investment cost. An artificial neural network
model was used to optimise a molecular distillation column in the palm
oil refinery process based on feed flow rate, column temperature and
pressure to optimise beta-carotene, tocopherol and free fatty acid con­
centration in the outlet stream [66]. Production of palm oil derived
cellulose phosphate is optimised by evaluating the reactor parameter
using wavelet neural network with a prediction error of less than 7%
[67]. Palm oil mill operating conditions were optimised via genetic al­
gorithm and artificial neural network to ensure the plant emissions are
within the allowable limit [68]. These high computational demanding
optimisation model via deep learning and artificial neural network can
easily be implemented by SMEs with cloud computing to minimise in­
vestment costs.
Other efforts and advantages of cloud computing also have been
exploited by researchers for its application in other sectors which can be
incorporated into the palm oil industry. For example, a study shows that
cloud computing performs well in solving the fuzzy problem of a trust-
based access control model, which can be applied in the palm oil in­
dustry to model trust issues among SMEs in plantations [69]. For
example, the model can be used in the planning of oil palm supply chain
based on the consistency and on-time delivery of oil palm from various
SMEs to avoid inadequacy of feedstock at the palm oil refinery. A
detailed review of cloud computing integration into supply chain system
shows improvement in efficiency and effectiveness, such as improved
communication and real-time information sharing [70]. Natural hazard
modelling system takes advantage of cloud computing as an easily
scalable computational power to handle increasing data collection from
various IoT detection system [71]. Similar applications can be incor­
porated with localised weather forecast devices in oil palm plantation to
compute local weather forecast and production yield using a cloud
computing platform, as discussed in Section 3.1. In order to reduce data
handling in cloud computing, a review of various scheduling approaches
shows that parameters such as response time, reliability, availability,
cost, and energy consumption are the key factors to implement cloud
computing [72]. Apart from the many applications of cloud computing,
researchers also focus on increasing the connection speed. For example,
edge and fog computing are developed to reduce the latency of the cloud
computing network [73]. These research efforts are essential to enhance
the feasibility of implementation of cloud computing in the palm oil
industry, especially for the upstream processes like plantation sites,
where internet accessibility and reliability could be the bottleneck of the
system.
3.5. Smart monitoring devices-based implementation
Advancement of Industry 4.0 technologies improves the traditional
monitoring system by reducing the requirement of manual operation
and labour. For instance, the features discussed in Sections 3.1–3.3
involved various IoT devices that used to detect the process parameters
and shared to the centralised system. Incorporation of IoT devices also
enables remote monitor and control of the process parameters. For
example, oil palm ripeness, palm tree health, water management at the
plantation site and process parameters in the palm oil refinery can be
accessed outside of work premises for better management efficiency.
Those suggestions are reflected in Recommendations 3 (AI imaging
processing technology to determine tree health status locally), 5 (AI
image processing technology to determine the ripeness of palm fruit
locally) and 17 (Incorporation of IoT devices in the process to enable
C.H. Lim et al.
remote monitoring and access for process control). Apart from the
process monitoring system, this section discusses the potential imple­
mentation of smart monitoring devices on personnel. From industry
feedback and discussion, several cases were found that employees left
their working position outside of the company compound during office
hours without consent from superior. This is critical in the oil palm
plantation due to the harvesting efficiency is heavily dependent on the
availability of labour. Global Positioning System (GPS) tracking function
from the company issued smartphone was used to trace staff locations
during an emergency, especially when the staff on duty cannot be
reached. However, there is a loophole of this system, where unethical
employees can intentionally leave their smartphone in their office while
they are physically away from their position. One of the alternative
options to improve the monitoring system is by means of smart wearable
devices such as smartwatches.
The current state of smartwatch technology not only can track the
user’s location via GPS but also incorporate healthcare functionalities
which can be used to monitor users’ health conditions. Oil palm farmers
are usually exposed to various working hazards such as hot weather that
may lead to dehydration, heavy lifting, potential injuries due to sharp
objects at plantation site or snake bite. This issue can be rectified with
biosensor embedded in a smartwatch to track the user’s activity from
heart rate assessment, movement detection and GPS information. For
example, operators’ heart rate can be monitored during operation in
plantation and processing plants to ensure good health status and
wearable sweat sensors can be incorporated to detect the physical state
of the wearer with health diagnostics [74,75]. Smartwatches were also
developed to detect falls, thus allowing first aid response teams to
achieve a swift response when the farmer or user has been determined to
have had a serious accident, especially during operation at height [76].
Apart from that, wearable biosensors have been slowly introduced in the
healthcare industry to provide a unique personalised healthcare plan to
suit individual needs based on their health conditions [77]. Tracing the
operators’ activity via biosensor in smartwatch will not only enhance
safety feature but can also be incorporated with insurance policy based
on the specific activity performed by operators with a robust tracking
system and record.
With the regular monitoring and recording, application of smart
monitoring system can be extended to various aspects. For example,
smart monitoring to improve product traceability which will be dis­
cussed in detail in Section 3.6. Besides, a smart contract between mul­
tiple stakeholders can be implemented, such as a dynamic incentive
reward and healthcare coverage system based on farmers daily working
activities tracked by wearable devices to identify scofflaws among their
peers. This would minimise the issue of employees intentionally remove
the smart wearable devices to temper with the tracking system as the
failure of tracking their activities could affect the incentive and
healthcare benefits. Further discussion on the smart contract application
will be presented in Section 3.7.
3.6. Traceability-based implementation
As discussed earlier, traceability of palm oil source is one of the key
challenges to promote CSPO due to the limitation of the current system
to differentiate sustainable and non-sustainable palm oil from the huge
amount of SMEs involved in the supply chain [8]. Besides, prolonged
storage of palm fruit and oil will affect the oil extraction yield. These
issues were addressed in the HAZOP Recommendations 9 (Blockchain
technology to enhance the traceability of products across the supply
chain) and 19 (Incorporate blockchain and IoT technology to ensure
each product is distributed within the maximum allowable storage
time). The recommendations suggested that the traceability of palm
oil-based products needs to be improved. In order to prove the palm oil is
sustainable, a robust tracking system that records the product lifecycle is
required to show the environmental impacts from each process. For
proper recording and tracking purposes, blockchain has shown an
8
Renewable and Sustainable Energy Reviews 135 (2021) 110223
excellent capability to process and broadcast various data along the
product lifecycle to enable information and service sharing across en­
terprises [78]. Blockchain is a shared, distributed, and synchronized
ledger that comprised of unchangeable and digitally recorded data in
blocks, which can facilitate transaction recording and assets tracking in
a business network [79]. This can be applied in the palm oil industry to
keep track of all palm oil-based products across each process to fulfil the
standard requirement, such as CSPO and HALAL certification. Study
shows that blockchain was used to improve supply chain management in
terms of traceability, fragmentation of the system, and capability to deal
with real-time information (sharing and control) [80].
For better security, blockchain technology should be embedded with
IoT devices such that the information captured by the IoT devices (such
as sensors) will be stored directly into the blockchain system to avoid
manual tempering by individuals. Researchers have conducted studies
to improved traceability of wearable IoT devices for healthcare service
to enhance privacy preservation, compromise detection and provide the
ability for audit tracing [81]. With the idea of incorporating IoT devices
to improve traceability as shown in Recommendation 12 (IoT devices
and GPS system to detect live loading weight and truck location during
distribution), an IoT weight sensor can be installed in a delivery truck to
trace the weight of palm oil products and the truck location. This
concern was raised during a discussion with oil palm plantation man­
ager where thievery is one of the potential issues in the supply chain of
fresh fruit bunches. Currently, all fresh palm fruits are weighted before
departure from the plantation site, and the weight is measured once
again at processing mill upon arrival to confirm the exact sales amount.
Often the amount is not tally due to potential losses during trans­
portation or thievery. The incorporation of IoT weight and GPS sensors
on delivery trucks may aid to minimise this problem. For example, if
weight losses are detected when the truck is in motion, most likely the
losses may occur due to transportation issues such as sharp corners,
bumpy road conditions or speeding. On the other hand, if the weight
losses are detected when the truck is in station mode or the truck loca­
tion deviates from the planned route based on GPS detection, an
investigation can be initiated for potential theft.
A similar type of application has been studied to improve traceability
by IoT devices. For instance, a radio-frequency identification system is
used in a container logistic system to identify different transport units
for monitoring and management purposes, improve efficiency, and
prevent container loss [82]. A study based on wood furniture supply
chain concluded that the investment cost for product tracking system
should be equitably distributed among all stakeholders to improve the
feasibility of implementation [83]. With such traceability capability
from Industry 4.0 technology, a more robust tracking system for CSPO
can help to tackle the issue of palm biodiesel production in the European
countries.
3.7. Smart contract-based implementation
A smart contract is the concept of an autonomous system that is
designed to conduct a transaction without human intervention. It is
usually coded parameters written into the blockchain to perform specific
transaction if certain requirement or condition is met prior (contractual)
agreement from all parties [84]. Smart contract applications are also
included in IoT, system security, and reduce the financial risk [85]. For
example, incorporation IoT and smart contract to enable automated
transactions or dedicated payment schemes. In line with the Recom­
mendation 4 (Health care devices such as smartwatches to monitor
health status and location of worker) in Table 2, a more efficient and fair
payment system can be programmed based on farmers’ health condition
and activity pattern in plantation to replace the conventional daily fix
rate scheme. Such implementation can be used to differentiate
committed employees from uncommitted employees and provide an
adequate reward (or penalty), respectively. From Recommendation 11
(Smart contract to determine the premium price for good quality
C.H. Lim et al.
products) and 16 (Incorporate smart contract to increase the selling
price of fresh palm fruit for SMEs with consistent supply), different
pricing of crude palm oil (CPO) based on its properties, quality, and
consistency of supply can be embedded into a smart contract to ensure a
consistent supply of premium palm oil, especially from SMEs. Farmers
that are able to produce premium palm oil that fulfils CSPO standard
will be automatically awarded the premium price in the smart contract
system to encourage SMEs to initiate CSPO certification and continue
with CSPO production.
Similar to the traceability function discussed in Section 3.6, a robust
smart contract system should be incorporated together with IoT-enabled
devices to allow for automatic data monitoring and recording to avoid
error from manual operation. For example, IoT-enabled containers are
utilised to monitor shipment conditions including temperature, location,
pressure (for open/close detection of a pressurised container), and ac­
celeration for sudden fall or drop detection [86]. If any of the sensors
detected violation of the agreed shipping condition, the information
could be used to adjust the shipping cost as per the smart contract agreed
by all stakeholders. This can be translated into the palm oil industry
where any unsustainable practice (e.g., deforestation which increases
carbon release) is detected, the premium price of CSPO will, therefore,
not be awarded. Application of smart contract in a blockchain-based
platform has shown promising results to improve the efficiency of
instant decision making and support for a better quality of product and
service [78]. Execution of transactions can be automated for an on-spec
product, and terminated if the product is founded to be faulty in any
process stage. Payment system for cloud services with smart contract
incorporation was evaluated in a study. Two models were proposed,
including a proof-based verification model which only feasible if all
parties are assumed to be honest; and a replication-based verification
model which works best when fines are imposed to cheating party [87].
The concept of Game Theory is incorporated into a model to mitigate
cheating and illegal behaviour to reduce the success rate of criminal
smart contract [88]. Another study discussed the application of smart
contract in an autonomous delivery system which includes consider­
ation of supply chain performance, legal aspects and product custom­
isation [89]. The smart contract also can be incorporated into financial
loan management where this technology can be beneficial to handle
subsidies or loan among SMEs [90]. For example, the loan interest rate is
imposed based on CSPO certification status to promote the initiation for
CSPO certification where a lower interest rate will be imposed for those
companies that are committed to fulfil the standard.
3.8. Cybersecurity-based implementation
Despite the advantages of applying Industry 4.0 technologies to
improve process efficiency, the system may be exposed to higher secu­
rity threat due to the interlinked nature of infrastructure. Cybersecurity
issues are related to several types of cyber-attacks, including viruses,
Trojan horse, worms, phishing, Denial-of-Service attack, illegal access
and control system attacks [91]. If any form of cybersecurity threat
manages to infiltrate the system of IoT devices, the rest of the
inter-linked devices are prone to the risk of sabotage as well. This
problem has been addressed in the HAZOP review in Recommendation
14 (Improve cybersecurity, such as blockchain for decentralised
network). Numerous researchers have proposed various methods to
evaluate and strengthen cybersecurity. A threat model is proposed to
conduct a risk analysis of an IoT system to assess cyber threats and
propose countermeasures [92]. Blockchain technology is incorporated
into the healthcare system to address the concern of information leakage
due to the decentralised nature of the blockchain system that resists to
cyber hack [93]. Various approaches have been discussed to improve
security issues in cloud, edge and fog computing platforms [94].
Blockchain technology can also be incorporated with smart contract
system to enhance operational security in battery energy storage system
from cyber-attack. The simulation result shows significant improvement
9
Renewable and Sustainable Energy Reviews 135 (2021) 110223
as compared to centralised or multi-agent distributed architectures [95].
A five-level trust model is proposed for cloud-edge based infrastructure
to allow data owners to determine the trust level of each party. This
increases the flexibility of the model to react differently in cases which
trust level can be adjusted based on specific action or performance [96].
As a robust and high-end cybersecurity system can be very expensive, a
study proposed an attacker-defender model to evaluate the strategy for
defending assets in a digital logistic network [97]. Similarly, Game
Theory has been applied to evaluate and optimise cloud security level
and its pricing strategy [98].
From the discussion, numerous efforts have been made by re­
searchers to ensure safe implementation of Industry 4.0 technologies.
Simultaneous implementation of cybersecurity and other Industry 4.0
technologies in the palm oil industry should be performed i) to avoid the
introduction of new forms of threat, and ii) to minimise operational risk
in future. However, based on industry input, the processing information
and knowledge in the palm oil industry are typically standard across
different companies. Less effort and investment have been used to
improve the security of information protection, except for the protection
of research and development information due to potential patent gen­
eration that may provide significant income to the company. Thus, any
form of cybersecurity measure should be given priority to the protection
of the research and development department during the transformation
of Industry 4.0 in the palm oil industry.
4. Potential Industry 4.0 implementation in the palm oil
industry
HAZOP approach is generally used in the chemical engineering in­
dustry to review and improve process safety based on P&IDs. A new
version of P&ID is produced to incorporate practical recommendations
from the HAZOP review to reflect the changes for design improvement.
By adopting the same philosophy, an updated block flow diagram (based
on Fig. 2) is proposed to reflect the potential Industry 4.0 technologies
implementation in the palm oil industry, as shown in Fig. 3. The pro­
posed potential implementations are based on the discussion of various
Industry 4.0 technologies in Sections 3.1–3.8. A total of 13 Industry 4.0
features have been proposed. These features include general features
that can be implemented across all stakeholders (Feature 1 to 3) and
specific features that can be adapted in particular process stages
(Feature 4 to 13). These features are proposed to tackle the current
challenges in the palm oil industry including traceability of CSPO,
inaccurate forecasting, poor labour and soil nutrient management, un­
planned process downtime, safety and healthcare [26,29,45,81,99].
Feature 1 (Smart contract for fair pricing) in Fig. 3 suggests that
blockchain technology can be implemented across all stakeholders, from
upstream palm oil processes to downstream market to improve product
lifecycle traceability and transaction security [78,93]. If blockchain
technology is implemented across all the palm oil industry stakeholders,
an industry-wide smart contract can be configured. An Industry 4.0
enabled industry can systematically generate fair transactions among
stakeholders based on their product sustainability performance, CSPO
certification, the carbon footprint from processes and logistics, and so­
cial impacts. Premium price or subsidy will be awarded or removed
depending on the traceable parameters. The improved traceability
function is critical to convince users, especially for palm biodiesel pro­
duction and usage in European countries, to show that the product
lifecycle is fulfilling the required sustainability standard. Feature 2
(Traceability of product losses) proposes a product tracing system for
logistics to avoid thievery, as discussed in Section 3.6. This will be
relatively easy to be implemented by a logistics company by installing
IoT-enabled sensors in delivery trucks or containers [82,83,89]. On the
other hand, the tedious digitalisation of printed documents creates
reluctancy for SMEs to adopt Industry 4.0 revolution. As such, the re­
view suggests using AI and deep learning technology in image process­
ing to digitalise document as presented by Feature 3 (Digitalisation of
C.H. Lim et al.
Fig. 3. Mapping of the potential implementation of Industry 4.0 technologies in the palm oil industry.
old documents) [58,100]. This feature is applicable to all stakeholders in
the palm oil industry as a supporting feature during the transition period
toward Industry 4.0.
Other features proposed from this study are more suitable to improve
specific palm oil processes, as shown on the right-hand side of Fig. 3. By
incorporating IoT devices and cloud computing at oil palm plantations,
the efficiency of plantation management can be improved. Feature 4
(Smart CPO production forecast) suggests that an enhanced forecast
system can be used for better yield prediction by installing IoT devices
for local weather; and soil detections to induce smart farming for better
water and fertiliser management, as suggested by Feature 5 (Smart
farming/irrigation) [27,36]. The data processing is proposed to be
conducted via cloud computing due to the consideration of relatively
low accessibility to powerful computing facilities at the plantation site.
Features 6 (Improve harvesting planning/forecast) and 7 (Optimum
harvesting route) are proposed to improve the efficiency of palm fruit
harvesting at the plantation site. The installation of IoT imaging devices
to detect the ripeness of palm fruit has been conducted by various re­
searchers [49–51]. By combining the IoT devices at plantation sites for
weather forecast and palm fruit ripeness detection, a more accurate
harvesting plan can be prepared to reduce labour cost as proposed in
Feature 6. A recent study suggested an inductive sensor system with up
to 100% accuracy in detecting palm fruit ripeness and an accurate es­
timate of harvest time [101]. This feature can be further improved by
incorporating AI and deep learning technology to optimise harvesting
route which aims to minimise the distance travelled by farmers within
the plantation site as proposed in Feature 7 (Optimum harvesting route)
[102,103]. Optimisation of palm fruit harvesting route is similar to
handling cold supply chain that the transported material’s quality will
deteriorate over time [104]. Feature 8 (Smart contract for price pri­
cing/salary) introduces wearable IoT devices with biosensors to monitor
the health conditions and activities of farmers at plantation sites [75,76,
105,106]. Such devices can be configured with smart contracts where
the workers are awarded and insured based on their daily performances
10
Renewable and Sustainable Energy Reviews 135 (2021) 110223
such as effectiveness in harvesting, duration of unplanned break time,
awareness to safety, and health status. This feature is not recommended
to be implemented in the downstream processing plants because palm
oil mill and refinery plant are usually operated in a more defined
boundary where supervision of workers is relatively easier as compared
to plantation sites.
Other Industry 4.0 technologies can be incorporated into down­
stream processes such as palm oil refineries to produce refined, bleached
and deodorised palm oil for palm biodiesel, food, and chemicals pro­
duction. Feature 9 (Dynamic process optimisation) proposes a dynamic
process optimisation and control system to improve the efficiency of the
processes. This requires the installation of IoT sensors in the processes
and computational facilities to analyse the dynamic fluctuation of big
data. For example, in biodiesel production from palm oil and waste
frying oil, the mixing ratio of the oils can be optimised based on the
dynamic changes of the feed properties and produced biodiesel prop­
erties [107]. Other alternative improvements include the optimisation
of the transesterification process in palm biodiesel production based on
methanol to oil ratio, amount of catalyst, reaction temperature, and
energy efficiency improvement [108]. With IoT sensors installed across
the processing system, two additional features can be configured as
stated in Feature 10 (Equipment defect detection) and 11 (Remote ac­
cess of process control). Feature 10 proposes a defect equipment
detection system to enable smart maintenance management to minimise
unplanned downtime [44,45]. A more advanced detection system may
utilise image processing technologies (e.g., photography and CCTV) to
determine any piping defect within the process plant, as discussed in
Section 3.3 [55]. Utilising the connection of IoT sensors with the control
system, remote access to the system can be configured as suggested in
Feature 11. This may reduce the requirement of senior engineer at the
palm oil processing plant, especially during the night shift. As a result,
the overall operation cost can be minimised; while enabling quick
remote respond during an emergency. Next, Features 12 (Smart restock
notification) and 13 (Smart production management) are proposed to
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223

install imaging processing and IoT devices to monitor the flow of
palm-based products at the retailer market. This aims to create a smart
marketing system where automated product replenishment can be
configured. In an ultimate scenario where the processing plants and
markets are connected with IoT systems, smart production planning can
be created to avoid oversupply or undersupply of palm-based product
and storage requirement can be minimised.
From the discussion, the majority of the proposed Industry 4.0 fea­
tures require the integration of IoT devices. Due to the interconnected
nature of IoT system, the enhancement of cybersecurity of the palm oil
industry should be given priority in the transformation. This should be
highlighted to protect information breaches such as research effort,
potential patent, personnel information (from wearable IoT devices),
CCTV footage, and process sabotage.
5. Conclusions
This paper implemented a HAZOP approach to review the current
state of the palm oil industry in Industry 4.0 revolution and develop­
ment. The state-of-the-art HAZOP review approach provides an in-depth
investigation of the system using a guidewords-based brainstorming
evaluation method. A total of 23 recommendations were proposed to
highlight the potential gaps to incorporate Industry 4.0 technologies in
the palm oil industry, including the use of IoT devices, smart image
processing technology, smart monitoring, smart forecasting, cloud
computing, traceability, smart contract, and cybersecurity. Analysis
from the review suggested that the combination of several Industry 4.0
technologies is required. For example, the combination of IoT devices
and cloud computing can enable smart CPO production forecast; while
the combination of IoT devices, image processing and cloud computing
can optimise harvest route in oil palm plantation. This study had iden­
tified a total of 13 potential Industry 4.0 features to be implemented
across all or specific palm oil industry stakeholders. A mapping diagram
was presented to show the possible improvement across the supply chain
of palm-based products. Several proposed features are beneficial for the
industry as a countermeasure to the movement of banning palm oil
usage in European countries due to the sustainability issue. Blockchain
technology from Industry 4.0 provides a secure transaction and data
recording to improve the traceability of the products. The imple­
mentation of such technology across palm oil industry stakeholders of­
fers a robust traceable system to develop sustainable profile and
certification of the product. This helps to promote CSPO production, and
consequently increase sustainable palm biodiesel production. Nonethe­
less, investment into IoT sensors for detection and data collection is
necessary for the Industry 4.0 transformation. In order to analyse the
massive amount of data from IoT devices, cloud computing was pro­
posed for big data analysis instead of physical computing facilities due to
almost none of the existing oil palm plantation and processing plant has
adequate computing capability for AI and big data analysis. The imaging
processing technology was proposed to improve the operational effi­
ciency, such as detection of fruit ripeness and faulty equipment.
In conclusion, this review paper has demonstrated the advantages of
HAZOP review approach in evaluating the status of the palm oil industry
in the revolution of Industry 4.0. This article discussed the current
problems and the conventional mitigation measures in the palm oil in­
dustry and explored the possible application of new Industry 4.0 tech­
nologies to improve the system. Nonetheless, the review was limited to a
macro-level of analysis for each stage of processes. A more detailed
consideration can be incorporated in future work. Instead of performing
a black box review of each process stage (e.g. the refinery), a more
detailed HAZOP literature review can be conducted to include various
process systems or equipment to provide a more holistic implementation
of Industry 4.0 features in the palm oil industry. Palm waste such as
palm biomass can also be included in a future study to understand the
overall footprint and the challenges toward palm oil industry 4.0.
CRediT authorship contribution statement
Chun Hsion Lim: Conceptualization, Methodology, Investigation,
Resources, Writing - original draft, Writing - review & editing, Visuali­
zation, Project administration, Funding acquisition. Steven Lim:
Investigation, Resources, Funding acquisition. Bing Shen How: Inves­
tigation, Resources, Writing - review & editing. Wendy Pei Qin Ng:
Investigation, Resources. Sue Lin Ngan: Investigation, Resources,
Writing - review & editing. Wei Dong Leong: Investigation, Resources.
Hon Loong Lam: Investigation, Resources.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.

Table A.1
HAZOP worksheet for Node 1
Node 1

Description: Palm oil upstream process including plantation, harvesting and milling

Deviation Guideword Causes Consequences Safeguards Recommendation

1. Feed 1. Fluctuation of 1. Uncertainty of weather 1. Harsh weather leading to
quantity affecting plantation yield low production yield.
and harvest activities.
2. Non-optimum water/ 1. Too much or too little supply
fertiliser usage. of water/nutrient resulting
lower yield of oil.
2. Over supply of water/
fertiliser resulting higher
unnecessary cost.
2. Inconsistency 1. Unhealthy palm tree. 1. Unhealthy palm tree leading
of quality to potential low production
yield and infection to other
trees.
2. Old tree. 1. Old tree causing low fruit
yield, resulting in non-
1. Regional weather prediction
model based on satellite data,
sea water level, sea buoys, etc.
1. Watering of plantation side
based on historical rain fall
quantity.
2. Fertiliser usage based on soil
and leaf analysis or
recommendation of fertiliser
supplier.
3. Scheduled annual fertiliser
application.
1. Satellite image to identify
unhealthy tree based on leaf
colour.
1. Scheduled replantation cycle
to ensure optimum yield.
1. Incorporation of localised
IoT devices to improve
the accuracy of weather
prediction.
2. Incorporation of localised
IoT devices to detect soil
properties for moisture
and nutrient.
3. AI imaging processing
technology to determine
tree health status locally.
(continued on next page)

11
C.H. Lim et al.
Table A.1 (continued )
Node 1
Description: Palm oil upstream process including plantation, harvesting and milling
Deviation Guideword Causes
3. Storage 1. Temporarily storage/piling
of fresh fruit at plantation
for transportation to mill
4. Various Not applicable
suppliers
2. Processing 1. Monitoring of 1. Lack of monitoring/
efficiency process supervision during
plantation and harvesting
process.
2. Harvesting based on manual
observation of fruit ripeness
in plantation side.
2. Process 1. Refer to fluctuation of
intervention/ water/fertiliser usage above
control that affect product quantity
and quality.
2. Complex harvesting plan
and scheduling of workers
across the plantation field.
3. Maintenance 1. Breakdown of tools,
machinery or vehicle.
3. Product 1. Fluctuation of 1. Less fresh palm fruit due to
quantity weather condition or
equipment breakdown.
2. More fresh palm fruits send
to milling plant which
exceeded the milling
capacity.
2. Inconsistency 1. Different plantation and
of quality harvesting practice among
farmers giving different
sustainability standard of
product (such as impact to
environment including
deforestation activity, water
usage, fertiliser impact).
2. Changes of physical and
chemical properties of
Consequences
optimal production rate vs
cost.
1. Potential prolonged piling of
fresh fruit with high
moisture leading to bacteria
growth, possible
contamination of fruit.
1. Inefficiency of farming/
harvesting activities
(possible leading to higher
labour cost).
2. Unable to detect unsafe
working environment for
workers
1. Manual observation (walk
around plantation) is labour
intensive activities which
may expose to human error
during observation, possible
miss-looked ripped fruits.
2. Ineffective forecast (based
on observation of fruit
ripeness) leading to
inefficient planning and
cost/resource management.
1. Quantity of ripped fruit from
each tree is unknow before
harvesting leading to
inefficient management of
resources resulting in higher
production cost.
1. Delay of process resulting
losses in production and off-
spec product.
1. Insufficient fresh fruit to
milling facility leading to
potential milling process
shutdown.
2. Insufficient CPO to
downstream processes
leading to possible delay of
production and process
shutdown.
1. Piling up of unprocessed
fresh fruit on milling plant
resulting in lower quality of
product (higher FFA
content), possible bacteria
growth.
1. Non-sustainable farming
practice leading to negative
impact to environment.
2. Difficulty to trace/ensure
sustainable practice among
small and medium-sized en­
terprise farmers.
1. Lower grade of fruit bunches
may potentially cause off-
12
Renewable and Sustainable Energy Reviews 135 (2021) 110223
Safeguards
1. Consistent and proper
management of supply chain,
such as all harvested fruits are
to be processed at mill within
6–24 h.
1. GPS tracking via smart phone
to locate personal (usually
limited to supervisor position
and above with company
phone).
1. Systematically segregate
harvesting zone based on
historical harvesting activity.
2. Prediction of fruit ripeness
based on previous
observation.
1. Workers/farmers are trained
to checked each tree for
ripped fruit within the
scheduled plantation area
(area which forecasted with
high number ripped fruit by
previous observation).
1. Scheduled monitoring and
checking of tools condition.
2. Schedule maintenance as per
vendor advise.
1. Multiple sources of fresh fruit
to each milling facility and
CPO to each process plant to
allow certain buffering time
for consistent process and
ensure security for required
supply volume.
1. Increase mill production time
to process all fresh fruits
within 24 h prior harvested.
2. Pretreatment of fresh fruit
(sterilize) to reduce
contamination.
1. Sustainable palm oil campaign
and certification such as
Malaysia Sustainable Palm Oil
(MSPO) to ensure high quality
of palm oil sources.
1. Pretreatment on all palm fruit
bunches are conducted to
Recommendation
4. Health care device such
as smartwatches to
monitor health status and
location of worker.
5. AI image processing
technology to determine
ripeness of palm fruit
locally.
6. Combined IoT device to
detect ripped fruit and
cloud computing to
optimise harvesting plan/
route to save time.
7. Incorporate IoT and big
data analysis on
machinery, detectors,
controllers to determine
optimum repair schedule
prior to breakdown.
8. IoT and big data analysis
to improve forecasting
techniques for fresh palm
fruit generation.
9. Blockchain technology
to enhance traceability
of products across the
supply chain.
10. Cloud computing to
optimise supply chain
network from multiple
source to achieve
required quantity,
properties, and
sustainability profile.
11. Smart contract to
determine the premium
(continued on next page)
C.H. Lim et al.
Table A.1 (continued )
Node 1
Description: Palm oil upstream process including plantation, harvesting and milling
Deviation Guideword Causes
harvested palm oil fruit
bunches (for example due to
weather condition).
3. Storage 1. Quality degradation of palm
oil fruit bunches due to
prolong storage.
4. Product 1. Unable to deliver product on
distribution time.
2. Unable to trace fruit losses
during transportation
4. Cost 1. Fluctuation of 1. Higher cost of organic
feed cost fertiliser compared to
inorganic fertiliser.
2. Fluctuation of 1. Dynamic price of CPO.
product price
3. Processing cost 1. Unplanned shutdown of
process.
5. Accuracy of 1. Traceability of 1. Manual documentation and
information resources recording method.
2. Information 1. Leak of existing process/
security technology information,
however this is considered
low impact as technology in
oil palm plantation and
harvesting is relatively
standard and well published.
2. Leak of research and
development information.
Consequences
spec/contamination of palm
oil.
2. Unfair costing (same selling
price across different quality
of fruit bunches).
1. Long storage promotes
bacteria growth, resulting
off-spec product.
1. Potential higher
transportation due to the
delayed deliver (such as
traffic, vehicle breakdown).
2. Prolonged delivery time
reduces quality of fresh palm
fruit (increased of FFA and
bacteria) leading to
potential off-spec product.
1. Loss of fresh palm fruit
during transportation (due
to insecure loading or
thievery) leading to profit
losses.
1. Potential high risk of
investment into new type of
fertiliser without justified
return.
1. Potential changes of palm
fruit bunches price leading
to unsecure or unfair pricing
from dealers.
2. Unable to differentiate
prices for different quality of
product
1. Emergency shutdown or less
feed scenario leading to
higher production cost to
maintain utilities supply,
especially boilers. However,
the effect is less significant
as majority of palm oil mill is
using EFB as boiler feed.
1. Potential loss of tracing of
each batch of product
resulting wrong
documentation and
recording.
1. Potential significant losses in
R&D, such as patent.
13
Renewable and Sustainable Energy Reviews 135 (2021) 110223
Safeguards Recommendation
minimise chances of off-spec price for good quality
product. products.
1. Pretreatment such as
sterilization in milling process
to minimise impact of
microorganism.
1. Location of plantation side
and mill is relatively close.
2. Use of GPS to detect road
condition.
3. Transport during off-peak
hours to avoid heavy traffic.
1. Truck loading weight check 12. IoT devices and GPS
prior departing from system to detect live
plantation and upon receival loading weight and
at mill to determine losses truck location during
during transportation. distribution.
1. Partial testing of new fertiliser
to compare and monitor the
improvement, such as oil yield
to minimise investment risk.
1. Forecasting of palm oil Previous Recommendations
production to estimate CPO 8 and 11 are applicable here:
price in future for planning 8. IoT and big data analysis
purpose. to improve forecasting
techniques for fresh
palm fruit generation.
11. Smart contract to
determine the premium
price for good quality
products.
1. Current practice to ensure
adequate amount of EFB to be
stored to maintain boiler
operation for specific duration
as per company operation
philosophy.
13. Provide tools for
digitalisation of printed
documents.
Previous Recommendation 9
is applicable here:
9. Blockchain technology
to enhance traceability
of products across the
supply chain.
1. Private and confidential 14. Improve cybersecurity,
contract with employees. such as blockchain for
2. Encrypted computer/laptop to decentralised network.
avoid information extraction
or sharing prior approval.
C.H. Lim et al.
Table A.2
HAZOP worksheet for Node 2
Node 2
Description: Processes of palm oil including refinery, bio-diesel production, and food/pharmaceutical production
Deviation Guideword Causes
1. Feed 1. Fluctuation of 1. Impact of regional CPO
quantity production yield such as
harsh weather, diseases,
seasonality yield etc.
2. Inconsistent supply from
various suppliers,
especially SMEs.
2. Inconsistency 1. Different physical and
of quality chemical properties of
CPO from different
suppliers.
2. Different sustainability
profile from different
suppliers due to
different farming,
harvesting practices
3. Storage 1. Prolonged storage of
CPO prior processing to
downstream product.
4. Various 1. Involvement of multiple
suppliers suppliers, especially
from SMEs.
2. Processing 1. Monitoring of 1. Visual inspection and
efficiency process monitoring of process
condition.
2. Process 1. Localised process
intervention/ control for each control
control loop.
2. No communication
between process control
loops.
3. Maintenance 1. Unplanned process
shutdown due to
equipment breakdown.
Consequences
1. Major impact to process,
potential process shutdown.
1. Potential shutdown of milling
process during low resource of
fruit bunches resulting higher
cost for starting-up and shut­
ting down the process.
2. Less impact to refinery and
downstream processes due
CPO has relatively longer shelf
life compared to fruit bunches
to allow storage for smooth
supply.
1. Potential product off-spec,
however this is very unlikely
due to establishment of
adequate pretreatment
processes.
1. Potential unable to achieve
MSPO requirement if the
source has high environmental
impact.
1. Potential off-spec of CPO
affecting efficiency of down­
stream processes, however this
is considered as unlikely event
due to CPO is relatively stable
for storage.
1. Potential inconsistency of raw
material quality, which has
same consequences as
“Inconsistency in quality” in
Node 2.
1. Localise inspection which may
cause potential exposure of
hazardous environment to
personnel.
2. Visual inspection may not able
to detect process deviation
accurately.
1. Potential cascade effect from
single control failure which
may affects multiple process
equipment and possible
process shutdown.
1. Unable to configure preventive
control between independent
control system/equipment.
1. Frequent unplanned shutdown
cause loss of production and
potential equipment damage.
14
Renewable and Sustainable Energy Reviews 135 (2021) 110223
Safeguards Recommendation
1. Storage system of CPO 15. Improvement of CPO
for consistent supply to production forecast for better
refinery and supply chain and storage
downstream processes. management.
1. Storage system of CPO 16. Incorporate smart contract to
for consistent supply to increase the selling price of
refinery and fresh palm fruit for SMEs with
downstream processes. consistent supply.
1. Palm oil source is Previous Recommendations 9 and
limited suppliers which 10 are applicable here:
have MSPO 9. Blockchain technology to
certification. enhance traceability of products
across the supply chain.
10. Cloud computing to optimise
supply chain network from
multiple sources; while all the
required quantity, properties,
and sustainability profile are
achieved.
1. Monitoring and control
of storage condition.
2. Various studies for
optimum palm oil
storage has been
conducted.
1. Provide adequate 17. Incorporation of IoT devices in
protection such as the process to enable remote
personnel protective monitoring and access for
equipment when process control.
conduct routine Previous Recommendation 7 is
monitoring at applicable here:
hazardous area. 7. Incorporate IoT and big data
2. Process monitoring and analysis on machinery,
control in control room. detectors, controllers to
determine optimum repair
schedule prior to breakdown.
1. Mitigation control Previous Recommendation 17 is
system for cascade applicable here:
effect, such as 17. Incorporation of IoT devices in
sequential unit the process to enable remote
shutdown due to monitoring and access for
process deviation. process control.
1. Current process control 18. Incorporation of big data
philosophy is based on analysis, cloud computing and
cascade control to IoT to enable communication
prevent further between multiple control loop/
escalation of process system for mitigation action
hazard. upon any control loop failure.
Previous Recommendation 7 is
applicable here:
7. Incorporate IoT and big data
analysis on machinery, detectors,
controllers to determine
(continued on next page)
C.H. Lim et al.
Table A.2 (continued )
Node 2
Description: Processes of palm oil including refinery, bio-diesel production, and food/pharmaceutical production
Deviation Guideword Causes
2. Complex maintenance
schedule due to huge
number of equipment in
each process.
3. Product 1. Fluctuation of 1. Major shortage of raw
quantity material.
2. Inconsistency 1. Unlikely to happen due
of quality to adequacy of process
control and quality
assurance & checking.
3. Storage 1. Over-storage of product
in warehouse.
4. Product 1. Complex logistic
distribution problem such as traffic
condition, weather,
scheduling, etc.
2. Required pre-booking/
order of product by
consumers/retailers
4. Cost 1. Fluctuation of 1. Fluctuation of CPO cost
feed cost based on the forecasted
production rate
2. Fluctuation of 1. Increase of CPO and
product price processing cost in
refinery.
2. Increase of treated palm
oil price and processing
cost in downstream
processes
3. Processing cost No significant impact
5. Accuracy of 1. Traceability of Same caused identified as
information resources in Node 2 and applicable
here.
2. Information Same caused identified as
security in Node 2 and applicable
here.
Consequences
1. Potential missing out
maintenance of major
equipment leading to
unplanned shutdown and
equipment damage.
1. Potential process disruption
and possible shutdown.
2. Unable to supply adequate
product to market/consumers,
affecting reputation and price.
1. Potential pro-longed storage of
product and exceeded recom­
mended shelf life.
2. Difficulty to trace different
batch of product.
1. Potential high logistic cost due
to non-optimum distribution
network, such as traffic jam.
2. Delay delivery due to weather,
traffic condition, incompetent
drivers.
1. Potential low or no product at
consumers/retailers’ side due
to unable to fulfill the required
amount or delayed delivery.
1. Inconsistency of raw material
price resulting in potential
higher production cost, leading
to possible stop of production
or less interest of investment
into palm oil industry
2. High cost of CPO resulting in
infeasible of bio-diesel produc­
tion as compared to petro-
diesel
1. Potential increase of treated
palm oil price leading to high
production cost.
1. Potential increase of
downstream product price.
However, this is considered as
unlikely event due to
downstream product has a
more significant value-added
profile where typical fluctua­
tion of palm oil price has min­
imum impact.
15
Renewable and Sustainable Energy Reviews 135 (2021) 110223
Safeguards Recommendation
optimum repair schedule prior to
breakdown.
1. Manual checklist for Previous Recommendations 7 and
maintenance. 13 are applicable here:
2. Scheduled preventive 7. Incorporate IoT and big data
maintenance analysis on machinery,
detectors, controllers to
determine optimum repair
schedule prior to breakdown.
13. Convert hardcopy of document
into softcopy via scanning
devices and software.
1. Product storage
management to allow
certain buffer time for
consistent supply.
19. Incorporate blockchain and IoT
technology to ensure each
product is distributed within
the maximum allowable storage
time.
1. Pre-fixed optimum 20. Incorporation of live weather
distribution network to forecast, traffic information and
minimise cost, such as GPS to optimise distribution
minimise transport network.
distance.
2. Schedule delivery
during off peak hours
(such as after night
time) to improve
efficiency.
21. Big data analysis to forecast
market demand for each
downstream product to plan
production strategy in advance.
22. Big data analysis to improve
CPO generation forecast
including weather and process
conditions to enable more
accurate prediction of pricing.
1. Adequate amount of Previous Recommendation 22 is
CPO storage for applicable here:
processing to minimise 22. Big data analysis to improve
impact of price CPO generation forecast
fluctuation. including weather and process
conditions to enable more
accurate prediction for pricing.
1. Adequate amount of
CPO storage for
processing to minimise
impact of price
fluctuation.
2. Secure more consistent
feed supply via multiple
suppliers, including
other oil crops sources.
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223

Table A.3
HAZOP worksheet for Node 3
Node 3

Description: Market for products/end users for palm oil product including palm bio-diesel, edible palm oil and palm chemicals products

Deviation Guideword Causes Consequences Safeguards Recommendation

1. Feed 1. Fluctuation of 1. Less production from process 1. Potential supply less than 1. Palm oil is one of the highest
quantity plant due to resources demand resulting temporary yield oil crops, with multiple
deficiency and unable to increase of product price. source from Malaysia,
supply to downstream 2. Prolonged insufficient Indonesia and other
market/consumers. product for market may loss countries with good security
consumers interest and trust, in supply chain. Thus, major
eventually less demand. deficiency of resources from
3. May encourage development multiple sources is unlikely
of new product or replaced by to happen.
alternative source which
affects overall palm oil
market.
2. Over production surpassing 1. Oversupply of product 1. Contract based supply chain Previous Recommendation
market demand. resulting in potential price system to ensure a minimum 21 is applicable here:
reduction of downstream transaction between 21. Big data analysis to
product, treated palm oil and stakeholders. forecast market
CPO, which possible affects 2. Current demand for palm oil demand for each
upstream processes, is relatively higher than the downstream product
especially to SMEs in supply capability, especially to plan production
plantation. in food and chemicals. strategy in advance.
2. Inconsistency 1. Unlikely to happened
of quality considering proper quality
assurance and checking in the
refinery and process plants.
3. Storage 1. Over storage of downstream 1. Prolonged storage of palm oil 1. Proper labeling of expiry
product. product resulting in product date of each palm oil product
storage time exceeded shelf on packaging.
life. 2. Treated and processed palm
oil product usually has
relatively long shelf life.
Thus, no significant impact.
4. Various 1. Not applicable assuming each
suppliers brand of product is unique
and coming from dedicated
process plant/company.
2. Processing 1. Monitoring of 1. Localised monitoring of 1. Manual and localised 1. Periodic stock count. 23. IoT and image
efficiency process product flow/sales in monitoring of product sales 2. Stock management system, processing devices to
respective market/retailer to and quantity may overlook such as replenish high keep track of stock
plan for stock replenishment. the timeline for stock demand product more amount daily.
replenishment resulting frequently and reduce order
temporary shortage of stock. amount for short expiry date
2. Process plant/supplier may product.
not able to provide immediate
supply for short notice order
resulting product shortage in
market.
2. Process Not applicable for end users/
intervention/ consumers
control
3. Maintenance Not applicable for end users/
consumers
3. Product 1. Fluctuation of 1. Increasing market demand. 1. Rapid increase of market 1. Unlikely event for sudden Previous Recommendation
quantity demand leading to increase of demand. 21 is applicable here:
insufficient supply of product, 2. Forecasting technique was 21. Big data analysis to
resulting insecure supply used to estimate the demand, forecast market
capability to end users/ such as bio-diesel. demand for each
customers. downstream product
to plan production
strategy in advance.
2. Decreasing market demand. 1. Lack of demand resulting Previous Recommendation
oversupply of palm oil 21 is applicable here:
product leading to price drop, 21. Big data analysis to
possible impact to palm oil forecast market
processing company and demand for each
upstream processes. downstream product
2. Refer to consequences in to plan production
“Fluctuation of quantity” in strategy in advance.
Node 3
2. Inconsistency Unlikely to happened
of quality considering proper quality
(continued on next page)

16
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223

Table A.3 (continued )

Node 3

Description: Market for products/end users for palm oil product including palm bio-diesel, edible palm oil and palm chemicals products

Deviation Guideword Causes Consequences Safeguards Recommendation

assurance and checking in the

refinery and process plants.
3. Storage Not applicable
4. Product Not applicable
distribution
4. Cost 6. Fluctuation of Same caused identified as
feed cost “Fluctuation of product price” in
Node 2 and applicable here.
7. Fluctuation of Not major issue of concern as the
product price selling price of palm oil product
will not have sudden changes.
8. Processing cost Not consider operation of retail
this review.
5. Accuracy of 1. Traceability of 1. Unable to trace the source of 1. Unable to trace each batch of 1. Enforce all stakeholders to Previous Recommendation
information resources product due to the complex product with respect to participant in product 9 is applicable here:
supply chain from plantation different batch of resources. standard certification, such 9. Blockchain technology
to end users, such as Halal 2. Unknown source of product as MSPO and HALAL to enhance traceability
certification for food or diminishes end users/ certificate. of products across the
sustainability profile/ consumers confident towards supply chain.
performance. palm oil product.
2. Information Not applicable.
security

[11] Higgins V, Richards C. Framing sustainability: alternative standards schemes for

Acknowledgement
Special appreciation to Mr Ezzaruddin Abdul Rapar from the Sime
Darby Plantation for the industry input; and financial support by Uni­
versiti Tunku Abdul Rahman Research Fund: IPSR/RMC/UTARRF/
2018-C2/L05.
References
[1] Shahnazi R, Shabani ZD. Do renewable energy production spillovers matter in the
EU? Renew Energy 2020;150:786–96. https://doi.org/10.1016/j.
renene.2019.12.123.
[2] Khatun R, Reza MIH, Moniruzzaman M, Yaakob Z. Sustainable oil palm industry:
the possibilities. Renew Sustain Energy Rev 2017;76:608–19. https://doi.org/
10.1016/j.rser.2017.03.077.
[3] Ahmad FB, Zhang Z, Doherty WOS, O’Hara IM. The outlook of the production of
advanced fuels and chemicals from integrated oil palm biomass biorefinery.
Renew Sustain Energy Rev 2019;109:386–411. https://doi.org/10.1016/j.
rser.2019.04.009.
[4] Purnomo H, Okarda B, Dermawan A, Ilham QP, Pacheco P, Nurfatriani F, et al.
Reconciling oil palm economic development and environmental conservation in
Indonesia: a value chain dynamic approach. For Policy Econ 2020;111:102089.
https://doi.org/10.1016/j.forpol.2020.102089.
[5] Sumathi S, Chai SP, Mohamed AR. Utilization of oil palm as a source of renewable
energy in Malaysia. Renew Sustain Energy Rev 2008;12:2404–21. https://doi.
org/10.1016/j.rser.2007.06.006.
[6] Awalludin MF, Sulaiman O, Hashim R, Nadhari WNAW. An overview of the oil
palm industry in Malaysia and its waste utilization through thermochemical
conversion, specifically via liquefaction. Renew Sustain Energy Rev 2015;50:
1469–84. https://doi.org/10.1016/j.rser.2015.05.085.
[7] Tey YS, Brindal M, Darham S, Ahmad SSF, Djama M. Early mover advantage in
Roundtable on Sustainable Palm Oil certification: a panel evidence of plantation
companies. J Clean Prod 2020;252:119775. https://doi.org/10.1016/j.
jclepro.2019.119775.
[8] Ruysschaert D, Salles D. Towards global voluntary standards: questioning the
effectiveness in attaining conservation goals the case of the Roundtable on
Sustainable Palm Oil (RSPO). Ecol Econ 2014;107:438–46. https://doi.org/
10.1016/j.ecolecon.2014.09.016.
[9] Gatti RC, Liang J, Velichevskaya A, Zhou M. Sustainable palm oil may not be so
sustainable. Sci Total Environ 2019;652:48–51. https://doi.org/10.1016/j.
scitotenv.2018.10.222.
[10] Murad SMA, Hashim H, Jusoh M, Zakaria ZY. Enhancement of roundtable on
sustainable palm oil (RSPO): quantitative assessment tool. Chem Eng Trans 2019;
72:67–72. https://doi.org/10.3303/CET1972012.
sustainable palm oil and South-South trade. J Rural Stud 2019;65:126–34.
https://doi.org/10.1016/j.jrurstud.2018.11.001.
[12] Tapia JFD, Samsatli S. Integrating fuzzy analytic hierarchy process into a multi-
objective optimisation model for planning sustainable oil palm value chains. Food
Bioprod Process 2019;119:48–74. https://doi.org/10.1016/j.fbp.2019.10.002.
[13] Jamaludin NF, Muis ZA, Hashim H. An integrated carbon footprint accounting
and sustainability index for palm oil mills. J Clean Prod 2019;225:496–509.
https://doi.org/10.1016/j.jclepro.2019.03.312.
[14] Hsion C, Wen W, Chuen Z, Qi J, Jun T, Shen B. Circular sustainability
optimisation model for diverse oil crops feedstock system via element targeting
approach. Chem Eng Trans 2019;76:1111–6. https://doi.org/10.3303/
CET1976186.
[15] Mayorga MA, Cadavid JG, Palacios OYS, Vargas J, González J, Narváez PC.
Production of renewable diesel by hydrotreating of palm oil with noble metallic
catalysts. Chem Eng Trans 2019;74:7–12. https://doi.org/10.3303/CET1974002.
[16] Silalertruksa T, Gheewala SH. Environmental sustainability assessment of palm
biodiesel production in Thailand Oil palm production. Energy 2012;43:306–14.
https://doi.org/10.1016/j.energy.2012.04.025.
[17] Yee KF, Tan KT, Abdullah AZ, Lee KT. Life cycle assessment of palm biodiesel:
revealing facts and benefits for sustainability. Appl Energy 2009;86:S189–96.
https://doi.org/10.1016/j.apenergy.2009.04.014.
[18] Yasin MHM, Mamat R, Najafi G, Ali OM, Yusop AF, Ali MH. Potentials of palm oil
as new feedstock oil for a global alternative fuel: a review. Renew Sustain Energy
Rev 2017;79:1034–49. https://doi.org/10.1016/j.rser.2017.05.186.
[19] Klepper G. Winners and losers from the EU’s proposed ban on palm oil. Eco-
Business; 2018. https://www.eco-business.com/opinion/winners-and-losers-fro
m-the-eus-proposed-ban-on-palm-oil/.
[20] MITI. National policy on industry 4.0, ministry of international trade and
industry. Ministry of International Trade and Industry; 2018.
[21] Culot G, Nassimbeni G, Orzes G, Sartor M. Behind the definition of Industry 4.0:
analysis and open questions. Elsevier B.V.; 2020. https://doi.org/10.1016/j.
ijpe.2020.107617.
[22] Ghobakhloo M. Industry 4.0, digitization, and opportunities for sustainability.
J Clean Prod 2020;252:119869. https://doi.org/10.1016/j.jclepro.2019.119869.
[23] Dunjó J, Fthenakis V, Vílchez JA, Arnaldos J. Hazard and operability (HAZOP)
analysis. A literature review. J Hazard Mater 2010;173:19–32. https://doi.org/
10.1016/j.jhazmat.2009.08.076.
[24] Lim CH, Lam HL, Ng WPQ. A novel HAZOP approach for literature review on
biomass supply chain optimisation model. Energy 2018;146:13–25. https://doi.
org/10.1016/j.energy.2017.11.033.
[25] Elsevier. ScienceDirect. https://www.sciencedirect.com/; 2020.
[26] Ibragimov A, Sidique SF, Tey YS. Productivity for sustainable growth in
Malaysian oil palm production: a system dynamics modeling approach. J Clean
Prod 2019;213:1051–62. https://doi.org/10.1016/j.jclepro.2018.12.113.
[27] Doeswijk TG, Keesman KJ. Adaptive weather forecasting using local
meteorological information. Biosyst Eng 2005;91:421–31. https://doi.org/
10.1016/j.biosystemseng.2005.05.013.
[28] Mehrkanoon S. Deep shared representation learning for weather elements
forecasting. Knowl Base Syst 2019;179:120–8. https://doi.org/10.1016/j.
knosys.2019.05.009.

17
C.H. Lim et al.
[29] Carron MP, Pierrat M, Snoeck D, Villenave C, Ribeyre F, Surhadi, et al. Temporal
variability in soil quality after organic residue application in mature oil palm
plantations. Soil Res 2015. https://doi.org/10.1071/SR14249 Temporal.
[30] Rhebergen T, Zingore S, Giller KE, Frimpong CA, Acheampong K, Ohipeni FT,
et al. Closing yield gaps in oil palm production systems in Ghana through Best
Management Practices. Eur J Agron 2020;115:126011. https://doi.org/10.1016/
j.eja.2020.126011.
[31] Rhebergen T, Fairhurst T, Giller KE, Zingore S. The influence of water and
nutrient management on oil palm yield trends on a large-scale plantation in
Ghana. Agric Water Manag 2019;221:377–87. https://doi.org/10.1016/j.
agwat.2019.05.003.
[32] Comte I, Colin F, Grünberger O, Follain S, Whalen JK, Caliman JP. Landscape-
scale assessment of soil response to long-term organic and mineral fertilizer
application in an industrial oil palm plantation, Indonesia. Agric Ecosyst Environ
2013;169:58–68. https://doi.org/10.1016/j.agee.2013.02.010.
[33] Pauli N, Donough C, Oberthur T, Cock J, Verdooren R, Rahmadsyah, et al.
Changes in soil quality indicators under oil palm plantations following
application of “best management practices” in a four-year field trial. Agric
Ecosyst Environ 2014;195:98–111. https://doi.org/10.1016/j.agee.2014.05.005.
[34] Boafo DK, Kraisornpornson B, Panphon S, Owusu BE, Amaniampong PN. Effect of
organic soil amendments on soil quality in oil palm production Daniel. Elsevier;
2019. https://doi.org/10.1016/j.apsoil.2019.09.008.
[35] Glaroudis D, Iossifides A, Chatzimisios P. Survey, comparison and research
challenges of IoT application protocols for smart farming. Comput Network 2020;
168:107037. https://doi.org/10.1016/j.comnet.2019.107037.
[36] Krishnan RS, Julie EG, Robinson YH, Raja S, Kumar R, Huy P, et al. Fuzzy logic
based smart irrigation system using internet of things. J Clean Prod 2020;252:
119902. https://doi.org/10.1016/j.jclepro.2019.119902.
[37] Khachatryan H, Suh DH, Xu W, Useche P, Dukes MD. Towards sustainable water
management: preferences and willingness to pay for smart landscape irrigation
technologies. Land Use Pol 2019;85:33–41. https://doi.org/10.1016/j.
landusepol.2019.03.014.
[38] Shi J, Wu X, Wang X, Zhang M, Han L, Zhang W, et al. Determining threshold
values for root-soil water weighted plant water de fi cit index based smart
irrigation. Agric Water Manag 2020;230:105979. https://doi.org/10.1016/j.
agwat.2019.105979.
[39] Wang J, Chen M, Zhou J, Li P. Data communication mechanism for greenhouse
environment monitoring and control: an agent-based IoT system. China
Agricultural University; 2019. https://doi.org/10.1016/j.inpa.2019.11.002.
[40] Lee I. The Internet of Things for enterprises: an ecosystem, architecture, and IoT
service business model. Internet Things 2019;7:100078. https://doi.org/
10.1016/j.iot.2019.100078.
[41] Zou Y, Yu L, Tso GKF, He K. Risk forecasting in the crude oil market: a multiscale
Convolutional Neural Network approach. Physica A 2020;541:123360. https://
doi.org/10.1016/j.physa.2019.123360.
[42] Ahmed K, Sachindra DA, Shahid S, Iqbal Z, Nawaz N. Multi-model ensemble
predictions of precipitation and temperature using machine learning algorithms.
Atmos Res 2020;236:104806. https://doi.org/10.1016/j.atmosres.2019.104806.
[43] Chapman R, Cook S, Donough C, Lim YL, Ho PVV, Lo KW, et al. Using Bayesian
networks to predict future yield functions with data from commercial oil palm
plantations: a proof of concept analysis. Comput Electron Agric 2018;151:
338–48. https://doi.org/10.1016/j.compag.2018.06.006.
[44] Kobbacy KAH. Application of artificial intelligence in maintenance modelling and
management. IFAC Proc Vol 2012;4:22–3. https://doi.org/10.3182/20121122-2-
ES-4026.00046. IFAC; 2012, pp. 54–9.
[45] Lee WJ, Wu H, Yun H, Kim H, Jun MBG, Sutherland JW. Predictive maintenance
of machine tool systems using artificial intelligence techniques applied to
machine condition data. Procedia CIRP 2018;80:506–11. https://doi.org/
10.1016/j.procir.2018.12.019.
[46] Zhang X, Chen MY, Wang MG, Ge YE, Stanley HE. A novel hybrid approach to
Baltic Dry Index forecasting based on a combined dynamic fluctuation network
and artificial intelligence method. Appl Math Comput 2019;361:499–516.
https://doi.org/10.1016/j.amc.2019.05.043.
[47] Villa-Henriksen A, Edwards GTC, Pesonen LA, Green O, Sørensen CAG. Internet of
Things in arable farming: implementation, applications, challenges and potential.
Biosyst Eng 2020;191:60–84. https://doi.org/10.1016/j.
biosystemseng.2019.12.013.
[48] Hormaza P, Fuquen EM, Romero HM. Phenology of the oil palm interspecific
hybrid Elaeis oleifera × Elaeis guineensis. Sci Agric 2012;69:275–80.
[49] Juman MA, Wong YW, Rajkumar RK, Goh LJ. A novel tree trunk detection
method for oil-palm plantation navigation Mohammed. Comput Electron Agric
2016;128:172–80. https://doi.org/10.1016/j.compag.2016.09.002.
[50] Rincón SM, Hormaza PA, Moreno LP, Prada F, Portillo DJ, García JA, et al. Use of
phenological stages of the fruits and physicochemical characteristics of the oil to
determine the optimal harvest time of oil palm interspecific OxG hybrid fruits.
Ind Crop Prod 2013;49:204–10. https://doi.org/10.1016/j.indcrop.2013.04.035.
[51] Makky M. A portable low-cost non-destructive ripeness inspection for oil palm
FFB. Agric Agric Sci Procedia 2016;9:230–40. https://doi.org/10.1016/j.
aaspro.2016.02.139.
[52] Silalahi DD, Reano CE, Lansigan FP, Panopio RG, Bantayan NC. Using genetic
algorithm neural network on near infrared spectral data for ripeness grading of oil
palm (Elaeis guineensis Jacq.) fresh fruit. Inf Process Agric 2016;3:252–61.
https://doi.org/10.1016/j.inpa.2016.10.001.
[53] Shabdin MK, Shariff ARM, Johari MNA, Saat NK, Abbas Z. A study on the oil palm
fresh fruit bunch (FFB) ripeness detection by using Hue, Saturation and Intensity
18
Renewable and Sustainable Energy Reviews 135 (2021) 110223
(HSI) approach. IOP Conf Ser Earth Environ Sci 2016;37:012039. https://doi.org/
10.1088/1755-1315/37/1/012039.
[54] Liu J, Wang X, Wang T. Classification of tree species and stock volume estimation
in ground forest images using Deep Learning. Comput Electron Agric 2019;166:
105012. https://doi.org/10.1016/j.compag.2019.105012.
[55] Yin X, Chen Y, Bouferguene A, Zaman H, Al-hussein M, Kurach L. A deep learning-
based framework for an automated defect detection system for sewer pipes.
Autom ConStruct 2020;109:102967. https://doi.org/10.1016/j.
autcon.2019.102967.
[56] Nturambirwe JFI, Opara UL. Machine learning applications to non-destructive
defect detection in horticultural products. Biosyst Eng 2019;189:60–83. https://
doi.org/10.1016/j.biosystemseng.2019.11.011.
[57] Kim DH, Lee S, Jeon J, Song BC. Real-time purchase behavior recognition system
based on deep learning-based object detection and tracking for an unmanned
product cabinet. Expert Syst Appl 2020;143:113063. https://doi.org/10.1016/j.
eswa.2019.113063.
[58] Cilia ND, Stefano C De, Fontanella F, Marrocco C, Molinara M, Freca AS Di. An
end-to-end deep learning system for medieval writer identification. Pattern
Recogn Lett 2020;129:137–43. https://doi.org/10.1016/j.patrec.2019.11.025.
[59] Dasgupta J, Bhattacharya K, Chanda B. A holistic approach for Off-line
handwritten cursive word recognition using directional feature based on Arnold
transform. Pattern Recogn Lett 2016;79:73–9. https://doi.org/10.1016/j.
patrec.2016.05.017.
[60] Zhang X, Bengio Y, Liu C. Online and offline handwritten Chinese character
recognition: a comprehensive study and new benchmark. Pattern Recogn 2017;
61:348–60. https://doi.org/10.1016/j.patcog.2016.08.005.
[61] Ali O, Osmanaj V. The role of government regulations in the adoption of cloud
computing: a case study of local government. Comput Law Secur Rev 2020;36:
105396. https://doi.org/10.1016/j.clsr.2020.105396.
[62] Sun P. Security and privacy protection in cloud computing: discussions and
challenges. J Netw Comput Appl 2020;160:102642. https://doi.org/10.1016/j.
jnca.2020.102642.
[63] Khayer A, Talukder S, Bao Y, Hossain N. Cloud computing adoption and its
impact on SMEs’ performance for cloud supported operations: a dual-stage
analytical approach. Technol Soc 2020;60:101225. https://doi.org/10.1016/j.
techsoc.2019.101225.
[64] Shaharum NSN, Shz M, Ghani WAWAK, Samsatli S, Al-Habshi MMA, Yusuf B. Oil
palm mapping over Peninsular Malaysia using Google Earth Engine and machine
learning algorithms. Remote Sens Appl Soc Environ 2020;17:100287. https://doi.
org/10.1016/j.rsase.2020.100287.
[65] Lee JSH, Wich S, Widayati A, Koh LP. Detecting industrial oil palm plantations on
Landsat images with Google Earth Engine. Remote Sens Appl Soc Environ 2016;4:
219–24. https://doi.org/10.1016/j.rsase.2016.11.003.
[66] Tehlah N, Kaewpradit P, Mujtaba IM. Artificial neural network based modeling
and optimization of refined palm oil process. Neurocomputing 2016;216:
489–501. https://doi.org/10.1016/j.neucom.2016.07.050.
[67] Wanrosli WD, Zainuddin Z, Ong P, Rohaizu R. Optimization of cellulose
phosphate synthesis from oil palm lignocellulosics using wavelet neural networks.
Ind Crop Prod 2013;50:611–7. https://doi.org/10.1016/j.indcrop.2013.08.048.
[68] Ahmad AL, Azid IA, Yusof AR, Seetharamu KN. Emission control in palm oil mills
using artificial neural network and genetic algorithm. Comput Chem Eng 2004;
28:2709–15. https://doi.org/10.1016/j.compchemeng.2004.07.034.
[69] Kesarwani A, Khilar PM. Development of trust based access control models using
fuzzy logic in cloud computing. King Saud University; 2019. https://doi.org/
10.1016/j.jksuci.2019.11.001.
[70] Novais L, Maqueira JM, Ortiz-Bas Á. A systematic literature review of cloud
computing use in supply chain integration. Comput Ind Eng 2019;129:296–314.
https://doi.org/10.1016/j.cie.2019.01.056.
[71] Ujjwal KC, Garg S, Hilton J, Aryal J, Forbes-smith N. Cloud Computing in natural
hazard modeling systems: current research trends and future directions. Int J
Disaster Risk Reduct 2019;38:101188. https://doi.org/10.1016/j.
ijdrr.2019.101188.
[72] Kumar M, Sharma SC, Goel A, Singh SP. A comprehensive survey for scheduling
techniques in cloud computing. J Netw Comput Appl 2019;143:1–33. https://doi.
org/10.1016/j.jnca.2019.06.006.
[73] Parikh S, Dave D, Patel R, Doshi N. Security and privacy issues in cloud, fog and
edge computing. Procedia Comput Sci 2019;160:734–9. https://doi.org/
10.1016/j.procs.2019.11.018.
[74] Jin X, Gu F, Niu J, Yu S, Ouyang Z. HRCal: an effective calibration system for
heart rate detection during exercising. J Netw Comput Appl 2019;136:1–10.
https://doi.org/10.1016/j.jnca.2019.03.007.
[75] Legner C, Kalwa U, Patel V, Chesmore A, Pandey S. Sweat sensing in the smart
wearables era: towards integrative, multifunctional and body-compliant
perspiration analysis Christopher. Sensors Actuators A Phys 2019;296:200–21.
https://doi.org/10.1016/j.sna.2019.07.020.
[76] Chen L, Li R, Zhang H, Tian L, Chen N. Intelligent fall detection method based on
accelerometer data from a wrist-worn smart watch. Measurement 2019;140:
215–26. https://doi.org/10.1016/j.measurement.2019.03.079.
[77] Papa A, Mital M, Pisano P, Giudice M Del. E-health and wellbeing monitoring
using smart healthcare devices: an empirical investigation. Technol Forecast Soc
Change 2020;153:119226. https://doi.org/10.1016/j.techfore.2018.02.018.
[78] Liu XL, Wang WM, Guo H, Barenji AV, Li Z, Huang GQ. Industrial blockchain
based framework for product lifecycle management in industry 4.0. Robot
Comput Integrated Manuf 2020;63:101897. https://doi.org/10.1016/j.
rcim.2019.101897.
C.H. Lim et al.
[79] Drljevic N, Aranda DA, Stantchev V. Perspectives on risks and standards that
affect the requirements engineering of blockchain technology. Comput Stand
Interfac 2020;69:103409. https://doi.org/10.1016/j.csi.2019.103409.
[80] Wang Z, Wang T, Hu H, Gong J, Ren X, Xiao Q. Blockchain-based framework for
improving supply chain traceability and information sharing in precast
construction. Autom ConStruct 2020;111:103063. https://doi.org/10.1016/j.
autcon.2019.103063.
[81] Lomotey RK, Pry J, Sriramoju S. Wearable IoT data stream traceability in a
distributed health information system. Pervasive Mob Comput 2017;40:692–707.
https://doi.org/10.1016/j.pmcj.2017.06.020.
[82] Madlenák R, Madlenáková L, Kolarovszká Z. System of management and
traceability of logisticitems through new technologies. Procedia - Soc Behav Sci
2016;230:128–35. https://doi.org/10.1016/j.sbspro.2016.09.016.
[83] Appelhanz S, Osburg V, Toporowski W, Schumann M. Traceability system for
capturing, processing and providing consumer-relevant information about wood
products: system solution and its economic feasibility. J Clean Prod 2016;110:
132–48. https://doi.org/10.1016/j.jclepro.2015.02.034.
[84] Guadamuz A. All watched over by machines of loving grace: a critical look at
smart contracts. Comput Law Secur Rev 2019;35:105338. https://doi.org/
10.1016/j.clsr.2019.105338.
[85] Zheng Z, Xie S, Dai H, Chen W, Chen X, Weng J, et al. An overview on smart
contracts: challenges , advances and platforms. Future Generat Comput Syst
2020;105:475–91. https://doi.org/10.1016/j.future.2019.12.019.
[86] Hasan H, Alhadhrami E, Aldhaheri A, Salah K, Jayaraman R. Smart contract-
based approach for efficient shipment management. Comput Ind Eng 2019;136:
149–59. https://doi.org/10.1016/j.cie.2019.07.022.
[87] Dorsala MR, Sastry VN, Chapram S. Fair payments for verifiable cloud services
using smart contracts. Comput Secur 2020;90:101712. https://doi.org/10.1016/
j.cose.2019.101712.
[88] Zhang L, Wang Y, Li F, Hu Y, Au MH. A game-theoretic method based on Q-
learning to invalidate criminal smart contracts. Inf Sci 2019;498:144–53. https://
doi.org/10.1016/j.ins.2019.05.061.
[89] Prause G. Smart Contracts for Smart Supply Chains. IFAC Pap n.d.;52:2501–2506.
https://doi.org/10.1016/j.ifacol.2019.11.582.
[90] Wang H, Guo C, Cheng S. LoC - a new financial loan management system based on
smart contracts. Future Generat Comput Syst 2019;100:648–55. https://doi.org/
10.1016/j.future.2019.05.040.
[91] Srinivas J, Das AK, Kumar N. Government regulations in cyber security:
framework, standards and recommendations. Future Generat Comput Syst 2019;
92:178–88. https://doi.org/10.1016/j.future.2018.09.063.
[92] Casola V, Benedictis A De, Rak M, Villano U. Toward the automation of threat
modeling and risk assessment in IoT systems. Internet Things J 2019;7:100056.
https://doi.org/10.1016/j.iot.2019.100056.
[93] Tripathi G, Ahad MA, Paiva S. S2HS- A blockchain based approach for smart
healthcare system. Elsevier Inc.; 2019. https://doi.org/10.1016/j.
hjdsi.2019.100391.
[94] Parikh S, Dave D, Patel R, Doshi N, Parikh S, Dave D, et al. Security and privacy
issues in cloud , fog and edge computing. Procedia Comput Sci 2019;160:734–9.
https://doi.org/10.1016/j.procs.2019.11.018.
19
Renewable and Sustainable Energy Reviews 135 (2021) 110223
[95] Mhaisen N, Fetais N, Massoud A. Secure smart contract-enabled control of battery
energy storage systems against cyber-attacks. Alexandria Eng J 2019;58:
1291–300. https://doi.org/10.1016/j.aej.2019.11.001.
[96] Chadwick DW, Fan W, Costantino G, Lemos R De, Di F, Herwono I, et al. A cloud-
edge based data security architecture for sharing and analysing cyber threat
information. Future Generat Comput Syst 2020;102:710–22. https://doi.org/
10.1016/j.future.2019.06.026.
[97] Cheung K, Bell MGH. Attacker – defender model against quantal response
adversaries for cyber security in logistics management : an introductory study.
Elsevier B.V.; 2019. https://doi.org/10.1016/j.ejor.2019.10.019.
[98] Kakkad V, Shah H, Patel R, Doshi N. A comparative study of applications of game
theory in cyber A comparative study of applications of game theory in cyber
security and cloud computing. Security and cloud computing. Procedia Comput
Sci 2019;155:680–5. https://doi.org/10.1016/j.procs.2019.08.097.
[99] Leegwater M, Duijn G van. 24 - traceability of RSPO-certified sustainable palm
oil. Palm Oil Prod. Process. Charact. Uses 2012:713–36.
[100] Greenbaum NR, Jernite Y, Halpern Y, Calder S, Nathanson LA, Sontag DA, et al.
Improving documentation of presenting problems in the emergency department
using a domain-specific ontology and machine learning-driven user interfaces. Int
J Med Inf 2019;132:103981. https://doi.org/10.1016/j.ijmedinf.2019.103981.
[101] Sinambela R, Mandang T, Subrata IDM, Hermawan W. Application of an
inductive sensor system for identifying ripeness and forecasting harvest time of oil
palm. Sci Hortic (Amst) 2020;265:109231. https://doi.org/10.1016/j.
scienta.2020.109231.
[102] He L, Xue M, Gu B. Internet-of-Things enabled supply chain planning and
coordination with big data services: certain theoretic implications. China Science
Publishing & Media Ltd.; 2020. https://doi.org/10.1016/j.jmse.2020.03.002.
[103] Lioutas ED, Charatsari C. Land Use Policy Smart farming and short food supply
chains: are they compatible? Land Use Pol 2020;94:104541. https://doi.org/
10.1016/j.landusepol.2020.104541.
[104] Theeb N Al, Smadi HJ, Al-Hawari TH, Aljarrah MH. Optimization of vehicle
routing with inventory allocation problems in Cold Supply Chain Logistics.
Comput Ind Eng 2020;142:106341. https://doi.org/10.1016/j.cie.2020.106341.
[105] Abouzahra M, Ghasemaghaei M. The antecedents and results of seniors’ use of
activity tracking wearable devices. Elsevier Ltd; 2019. https://doi.org/10.1016/j.
hlpt.2019.11.002.
[106] Cosoli G, Spinsante S, Scalise L. Wrist-worn and chest-strap wearable devices:
systematic review on accuracy and metrological characteristics. Measurement
2020;159:107789. https://doi.org/10.1016/j.measurement.2020.107789.
[107] Almeida VF de, García-Moreno PJ, Guadix A, Guadix EM. Biodiesel production
from mixtures of waste fish oil, palm oil and waste frying oil: optimization of fuel
properties. Fuel Process Technol 2015;133:152–60. https://doi.org/10.1016/j.
fuproc.2015.01.041.
[108] Khoobbakht G, Kheiralipour K, Rasouli H, Rafiee M, Hadipour M, Karimi M.
Experimental exergy analysis of transesterification in biodiesel production.
Energy 2020;196:117092. https://doi.org/10.1016/j.energy.2020.117092.