Professional Documents
Culture Documents
A Review of Industry 4.0 Revolution Potencial in A Sustainable and Renewable Palm Oil Industry
A Review of Industry 4.0 Revolution Potencial in A Sustainable and Renewable Palm Oil Industry
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.
oil, chemical product and energy, palm oil industry has been growing
rapidly due to the increasing demand. Global palm oil production in
1. Introduction creases tremendously in the past decades, almost doubling every 10
years [2]. Reported global industry consumption of palm oil and palm
The rapid development of industrialisation to improve the quality of kernel oil was about 19.72 million tons in 2018, while the highest
life, especially to meet the increasing demand of energy consumption producing country, Indonesia, alone aims to produce 40 million tons by
has significantly increased pollution. Recent developments focus on 2020 [3,4]. A study shows that palm oil is an energy crop which has the
reducing the dependence on fossil fuel by replacing it with other highest oil production yield as compared to other oil crops [5]. This
environmental-friendly and renewable resources such as solar, wind and enables optimum usage of limited plantation site (or land) to generate
geothermal. Significant research has been conducted to improve the highest amount of oil for food, chemical and energy applications.
renewable energy production to achieve the sustainable goal initiated by Apart from that, wastes generated from palm oil industry such as palm
the European Union to reduce 20% of the carbon dioxide and energy biomass and palm oil mill effluent have shown to be a promising
consumption by 2020 [1]. Being one of the primary sources for edible
* 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. Renewable and Sustainable Energy Reviews 135 (2021) 110223
2
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
literature review approach designed for this paper. processing, cloud computing, oil palm ripeness, blockchain, cloud
As per the first step in the procedure, Fig. 2 represents a simplified computing, smart contract, cybersecurity, big data analysis, and opti
block flow diagram of the palm oil supply chain used for this study. The misation. Recommendations were suggested to highlight the possible
overall process was classified into 3 nodes to enable a focused review integration of Industry 4.0 technologies into the palm oil industry as
process. The deviations and guidewords that used to trigger imagination research and development gaps. These recommendations can be adop
and brainstorming of potential process deviation/hazard in this study ted by stakeholders as a guideline to study the possible Industry 4.0
are summarised in Table 1. Several assumptions and limitations have implementation and investment for their respective sectors.
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 3. Results and discussion
technologies (implemented in existing palm oil and other industries, and
developing technologies) are considered in this study, and iii) any form A comprehensive HAZOP review was conducted based on the scope
of input (literature, interview, survey, etc.) from industry stakeholders presented in Fig. 2. The details of the discussions were recorded and
are considered as credible source and information. The HAZOP review documented in Table A.1, A.2 and A.3 in the appendices. Table 2 sum
was conducted to the best knowledge of authors based on available in maries all the recommendations proposed from the HAZOP review.
formation from literature and input from palm oil industry stakeholders. These recommendations are the compiled information for potential
Deviations and guidewords are used to trigger a structured imagination improvement and implementation of Industry 4.0 technologies into the
of possible process upsets or problems. Credible causes and their palm oil industry based on the current industry problems and maturity
respective ultimate consequences are also discussed and summarised in of Industry 4.0 technologies. A total of 14 recommendations were
worksheets (see Appendix). Upon assessing the problems and conse identified from the upstream processes (Node 1), 8 new recommenda
quences in the industry, the team also conducted research based on tions proposed for downstream processes (Node 2) and 1 new recom
literature and industry input to identify existing practices or methods mendation was introduced from market/retailers’ point of view (Node
that could mitigate/safeguard the problems. The identified problems 3). Analysis of the outcomes suggested that some recommendations
were related to existing Industry 4.0 technologies and features (that were interrelated and addressing a similar issue in the palm oil industry,
have yet to be implemented in the palm oil industry) to access the po which can be rectified or improved with Industry 4.0 technologies. A
tential adaptation which may help to minimise/prevent the problems. total of 8 classes of Industry 4.0 technologies were identified that have
The verification of existing technologies was mainly focused on the the potential to be implemented in the palm oil industry. The following
publications (within 5 years) from the Elsevier database, while some sections discussed the details of each technology and its respective
older publications were included if the content is highly relevant to this function in the palm oil industry.
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
3
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
3.1. IoT devices-based implementation subsequently reduces the oil extraction and profit [26]. Various re
searches were conducted to enhance the local weather prediction model.
Upon analysis and investigation, Recommendations 1 and 2 (sum For instance, an adaptive weather forecasting device based on local
marised in Table 2) proposed from the HAZOP review can be summar temperature, wind speed and global radiation was proposed for agri
ised as potential gaps to utilise IoT devices to improve oil palm culture application [27]. The proposed model has shown better accuracy
plantation management. Recommendation 1 (Incorporation of localised in providing short-term forecasts using local devices, including predic
IoT devices to improve the accuracy of weather prediction) suggested tion of seasonal changes based on weather conditions in previous years.
that IoT devices may help to minimise the impact on production yield Convolutional neural network architecture which capable of conducting
and activities due to uncertain weather conditions. Based on the oil palm accurate weather forecasts based on the collected regional temperature
plantation expert’s comments, harvesting of fresh palm fruit is highly and wind speed to improve the prediction performance [28]. Literature
dependent on weather conditions. Currently, daily harvesting activity is has shown that IoT devices can be implemented into upstream of the
planned according to the regional or national weather forecast, which palm oil industry to have a better local weather forecast system for more
does not reflect the actual condition at the plantation site. Thus, local stable harvesting and production planning.
ised devices to detect plantation site parameters for an improved Similarly, Recommendation 2 (Incorporation of localised IoT devices
weather prediction system may help to increase the efficiency of plan to detect soil properties for moisture and nutrient) addressed the need
tation management. A study has shown poor labour planning due to for local devices to detect and control soil properties, such as nutrients
inaccurate weather forecast may cause a delay in harvesting, thus and moisture for optimum plantation conditions. Soil nutrients may be
4
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. Renewable and Sustainable Energy Reviews 135 (2021) 110223
storage management), 21 (Big data analysis to forecast market demand improve the efficiency of processes in the palm oil industry. In the
for each downstream product to plan for production strategy in plantation site, identification of palm tree health status and fruit ripe
advance), and 22 (Big data analysis to improve CPO generation forecast ness has been one of the major process bottlenecks which involve
including weather and process conditions to enable more accurate pre intensive labour and time. Based on oil palm plantation expert, satellite
diction for pricing) in Table 2. As discussed in the previous section, the imaging technology can be used to identify unhealthy trees (based on
dynamic nature of palm oil production is directly affecting the pricing leaf colour) within a large plantation site. Nonetheless, local observation
and downstream applications, especially the competitive price com is still required for confirmation, especially for the case where a single
parison between biodiesel and petrodiesel. Therefore, the accuracy of (or a small number of) unhealthy tree is surrounded by many healthy
forecasting technology to predict the dynamic production and price of trees. Similar smart imaging technology can be implemented for palm
palm oil is crucial for the industry to have more precise management and fruit ripeness detection. Variation of flowering and ripeness of each palm
planning. A series of Industry 4.0-based forecasting technologies are fruit in the same brunch or tree is very common due to the non-
discussed in this section, e.g., the incorporation of big data analysis and homogenous pollination nature of oil palm [48]. This requires farmers
deep machine learning. Study shows that a convolutional neural to identify ripped fruits visually at the site for harvesting purpose. Such
network model can improve forecasting accuracy due to its ability to requirement and practice demand tremendous labour force; and poses a
capture and utilise multi-scale data characteristics [41]. Various ma relatively low efficiency. One of the potential adaptations of imaging
chine learning algorithms based on precipitation, maximum and mini processing in an oil palm plantation is to use localised imaging system
mum temperature data were applied to model climate changes in such as CCTV or mobile imaging devices (mobile robotic or drone) for
Pakistan including artificial neural network, K-nearest neighbour, sup the oil palm ripeness detection. The system should consider the uneven
port vector machine and relevance vector machine [42]. An accurate terrain and lighting at the plantation site. However, to the authors’ best
weather forecast provides a better basis of water and fertiliser man knowledge, such a system has yet to be implemented in the palm oil
agement, and hence improves the production yield of palm oil. Re industry.
searchers adapted Bayesian networks to predict oil palm plantation Despite that, on-going research and development to integrate smart
yield, which shows comparable results to the artificial neural network image processing technologies into oil palm process had been conduct
approach [43]. ed. A mobile robot was built to scan and identify palm tree trunk in
Forecasting technology can also be applied in the process control and plantation sites using colour image processing and depth data processing
management in the palm oil industry, such as palm oil mill and biodiesel techniques [49]. Detection of palm fruit ripeness by the red, green and
refinery. The need for a preventive measure in the processing was blue (RGB) colour unit from the image captured was investigated [50].
highlighted in Recommendation 18 (Incorporation of big data analysis, Similarly, a low-cost oil palm ripeness detector was constructed using a
cloud computing and IoT to enable communication between multiple digital camera and laptop to detect RGB of palm fruit placed in a small
control loop/system for mitigation action upon any control loop failure). chamber [51]. On the other hand, the incorporation of smart imaging
A comprehensive review was conducted on various AI applications in detection and computational power is required to accurately detect and
the management of process maintenance [44]. The study also showed predict ripeness of palm fruit for harvesting planning. Study shows that
that the genetic algorithm and neural network are the best approaches palm fruit ripeness can be detected with genetic algorithm neural
for preventive maintenance planning and scheduling purpose. Another network based on near-infrared spectral data [52]. An artificial neural
work was conducted to predict the requirement of maintenance on network model was developed to detect ripeness of fresh fruit bunch via
machine tools using AI techniques based on the machine condition data hue, saturation and intensity colour model. The result shows that the
to reduce unplanned downtime or failure [45]. This technology can be accuracy of the artificial neural network model is higher as compared to
implemented into the palm oil processes, equipment maintenance the linear regression model [53]. A deep learning model to classify tree
detection and scheduling system to minimise equipment breakage. For species can be modified to classify palm trees (with fruit and without
example, detection of mechanical failure in the oil extraction process fruit) for harvesting purpose [54]. Further improvement of the smart
based on the movement of the pressing machine and prediction of the imaging processing technology can be achieved by integrating the
maintenance schedule for the process control valve in refinery based on detection system with IoT and cloud computing as proposed in
the response time degradation. In terms of supply chain and logistics Recommendation 6 (Combined IoT device to detect ripped fruit and
management, researchers have improved the Baltic Dry Index fore cloud computing to optimise harvesting plan/route to save time). Un
casting accuracy with AI method to minimise risk due to changes in healthy trees or palm fruit ripeness can be detected with IoT-enabled
shipping freights [46]. detection robot or drone where the information can be processed in a
Implementation of Industry 4.0 technologies such as big data anal centralised cloud computing server for analysis. For instance, the smart
ysis, neural network and AI have shown the potential to improve the detection system can be used to propose an optimum harvesting route
accuracy in forecasting technology. Various researches have shown that for farmers to reduce the harvesting period or to recommend an opti
adaptation of such technology is possible in plantation, process, and mum schedule for the tree health check.
logistics. Nonetheless, for such technology to be integrated into the palm Apart from oil palm plantation site, the implementation of image
oil industry, significant improvement of the existing equipment and processing technology also can be applied in palm oil processing such as
facilities are required. For instance, IoT-enabled sensors have to be biodiesel refinery. One of the applications includes detection of defects
installed in plantation sites and processing plants for big data collection on process equipment and product to improve the efficiency of main
and input into the forecasting technology. This may impose challenges tenance activities and quality checking processes. Research has suc
for stakeholders, especially SMEs, in terms of affordability due to cessfully developed a deep learning model to detect piping defect via
various uncertainties in revenue and cost [47]. CCTV footage [55]. Implementation of such a detection system may help
to prevent unpredicted piping failure, especially for hazardous process
3.3. Smart image processing technology-based implementation piping and equipment. A comprehensive review of the application of
non-destructive sensors to detect horticultural products and the appli
From Recommendations 3 (AI imaging processing technology to cation of machine learning was conducted as part of the detection tools
determine tree health status locally), 5 (AI image processing technology to improve the accuracy of the system [56]. Image processing technol
to determine the ripeness of palm fruit locally), 13 (Provide tools for the ogy can also be applied in the retailer operation as suggested from
digitalisation of printed documents), and 23 (IoT and image processing HAZOP Recommendation 23 (IoT and image processing devices to keep
devices to keep track of daily stock count) in Table 2, the HAZOP review track of daily stock count). For instance, unmanned retailer shop or
shows that the advancement of image processing tool may help to stock-check process using CCTV and sensor to detect changes of palm oil
6
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
product on the shelve. This idea has been demonstrated in a study of across many industries and problems. A study was conducted by using a
real-time recognition system to detect and track products in the cabinet free cloud computing platform and image processing technology from
[57]. Complex smart tracking technology can further be improved by Google Earth Engine to classify the oil palm plantation map into water,
incorporation of demand forecasting based on the product flow detec built-up, forest, bare soil, oil palm, paddy and other vegetation [64].
tion on the shelve to manage stock replenishment accurately. For Nonetheless, distinguishing oil palm plantation with other vegetation
example, the production rate of palm biodiesel is planned according to from Google Earth image has proven to be a challenging work due to the
the demand of biodiesel to avoid over or under production to maintain a similarity in spectral image and structure [65]. Complex process opti
healthy biodiesel price. misation in palm oil processing plants can be solved with cloud
Development of image processing technology helps in the digital computing to reduce the investment cost. An artificial neural network
transformation of the palm oil industry and its operations. Information model was used to optimise a molecular distillation column in the palm
transfer and documentation such as specification sheet (or datasheet) of oil refinery process based on feed flow rate, column temperature and
palm oil product, checklist for maintenance, etc., are usually in hard pressure to optimise beta-carotene, tocopherol and free fatty acid con
copies, especially among SMEs. One of the reasons that many stake centration in the outlet stream [66]. Production of palm oil derived
holders are reluctant to put extra effort into Industry 4.0 is due to the cellulose phosphate is optimised by evaluating the reactor parameter
tedious transition period, such as digitalisation of system that includes using wavelet neural network with a prediction error of less than 7%
converting historical data from hardcopy into softcopy. This creates an [67]. Palm oil mill operating conditions were optimised via genetic al
additional barrier for the advancement of the palm oil industry into the gorithm and artificial neural network to ensure the plant emissions are
Industry 4.0 era. The issue was highlighted in Recommendations 13 within the allowable limit [68]. These high computational demanding
(Provide tools for the digitalisation of printed documents) where a smart optimisation model via deep learning and artificial neural network can
system is needed to aid the transformation towards the paperless system. easily be implemented by SMEs with cloud computing to minimise in
This paper has identified several researches to improve the efficiency of vestment costs.
translating hardcopy documents into softcopy via image processing. For Other efforts and advantages of cloud computing also have been
example, a deep learning system was constructed to detect row of sen exploited by researchers for its application in other sectors which can be
tences within a written document for better digital translation results incorporated into the palm oil industry. For example, a study shows that
[58]. Similar work was conducted to recognise different languages to cloud computing performs well in solving the fuzzy problem of a trust-
cater to different nationalities of workers. For example, English hand based access control model, which can be applied in the palm oil in
written word image identification was conducted based on Arnold dustry to model trust issues among SMEs in plantations [69]. For
transform approach to recognise and record written documents in soft example, the model can be used in the planning of oil palm supply chain
copy, while similar work was undertaken to recognise written Chinese based on the consistency and on-time delivery of oil palm from various
characters [59,60]. Adaptation of such technology may help to promote SMEs to avoid inadequacy of feedstock at the palm oil refinery. A
more initiatives from palm oil industry stakeholders to move towards detailed review of cloud computing integration into supply chain system
Industry 4.0. Nevertheless, the system should uphold a proper security shows improvement in efficiency and effectiveness, such as improved
system to avoid leakage of sensitive information. This issue will be communication and real-time information sharing [70]. Natural hazard
addressed with cybersecurity in Section 3.8. modelling system takes advantage of cloud computing as an easily
scalable computational power to handle increasing data collection from
3.4. Cloud computing-based implementation various IoT detection system [71]. Similar applications can be incor
porated with localised weather forecast devices in oil palm plantation to
Previous sections had discussed various potential implementations of compute local weather forecast and production yield using a cloud
Industry 4.0 technologies in the palm oil industry, including localised computing platform, as discussed in Section 3.1. In order to reduce data
IoT devices, forecasting and smart image processing. The operation of handling in cloud computing, a review of various scheduling approaches
such technologies requires a considerable amount of computing power shows that parameters such as response time, reliability, availability,
for big data analysis and AI processing. This creates a challenge among cost, and energy consumption are the key factors to implement cloud
palm oil industry stakeholders, especially SMEs, which do not have computing [72]. Apart from the many applications of cloud computing,
advanced computing facilities in their conventional operation. Cloud researchers also focus on increasing the connection speed. For example,
computing technology is one of the Industry 4.0 technologies which edge and fog computing are developed to reduce the latency of the cloud
enables multiple users to access a centralised powerful computing ma computing network [73]. These research efforts are essential to enhance
chine with dynamic and reconfigurable computing applications. The the feasibility of implementation of cloud computing in the palm oil
applications include remote access for information sharing, editing and industry, especially for the upstream processes like plantation sites,
storage, database management, and processing large datasets for com where internet accessibility and reliability could be the bottleneck of the
plex problems [61]. There are various types of cloud computing, system.
including private cloud, public cloud, community cloud and hybrid
cloud with their respective features and applications [62]. The main 3.5. Smart monitoring devices-based implementation
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 Advancement of Industry 4.0 technologies improves the traditional
processed result is delivered back to the users. This enables users to monitoring system by reducing the requirement of manual operation
access powerful computing facility with a minimum cost of subscription and labour. For instance, the features discussed in Sections 3.1–3.3
for the service to avoid huge investment cost into physical computing involved various IoT devices that used to detect the process parameters
facility. This feature could be a more favourable option to the palm oil and shared to the centralised system. Incorporation of IoT devices also
industry, especially SMEs in the plantation. A study has shown that enables remote monitor and control of the process parameters. For
migration to the cloud server for data storage and computing helps the example, oil palm ripeness, palm tree health, water management at the
organisation by reducing capital and operational cost, improve flexi plantation site and process parameters in the palm oil refinery can be
bility, and increase revenue, especially among SMEs [63]. Nonetheless, accessed outside of work premises for better management efficiency.
accessibility of cloud computing depends on internet access and con Those suggestions are reflected in Recommendations 3 (AI imaging
nectivity, which is a common facility readily available in the palm oil processing technology to determine tree health status locally), 5 (AI
industry, except in rural plantation sites. image processing technology to determine the ripeness of palm fruit
Various applications of cloud computing have been implemented locally) and 17 (Incorporation of IoT devices in the process to enable
7
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
remote monitoring and access for process control). Apart from the excellent capability to process and broadcast various data along the
process monitoring system, this section discusses the potential imple product lifecycle to enable information and service sharing across en
mentation of smart monitoring devices on personnel. From industry terprises [78]. Blockchain is a shared, distributed, and synchronized
feedback and discussion, several cases were found that employees left ledger that comprised of unchangeable and digitally recorded data in
their working position outside of the company compound during office blocks, which can facilitate transaction recording and assets tracking in
hours without consent from superior. This is critical in the oil palm a business network [79]. This can be applied in the palm oil industry to
plantation due to the harvesting efficiency is heavily dependent on the keep track of all palm oil-based products across each process to fulfil the
availability of labour. Global Positioning System (GPS) tracking function standard requirement, such as CSPO and HALAL certification. Study
from the company issued smartphone was used to trace staff locations shows that blockchain was used to improve supply chain management in
during an emergency, especially when the staff on duty cannot be terms of traceability, fragmentation of the system, and capability to deal
reached. However, there is a loophole of this system, where unethical with real-time information (sharing and control) [80].
employees can intentionally leave their smartphone in their office while For better security, blockchain technology should be embedded with
they are physically away from their position. One of the alternative IoT devices such that the information captured by the IoT devices (such
options to improve the monitoring system is by means of smart wearable as sensors) will be stored directly into the blockchain system to avoid
devices such as smartwatches. manual tempering by individuals. Researchers have conducted studies
The current state of smartwatch technology not only can track the to improved traceability of wearable IoT devices for healthcare service
user’s location via GPS but also incorporate healthcare functionalities to enhance privacy preservation, compromise detection and provide the
which can be used to monitor users’ health conditions. Oil palm farmers ability for audit tracing [81]. With the idea of incorporating IoT devices
are usually exposed to various working hazards such as hot weather that to improve traceability as shown in Recommendation 12 (IoT devices
may lead to dehydration, heavy lifting, potential injuries due to sharp and GPS system to detect live loading weight and truck location during
objects at plantation site or snake bite. This issue can be rectified with distribution), an IoT weight sensor can be installed in a delivery truck to
biosensor embedded in a smartwatch to track the user’s activity from trace the weight of palm oil products and the truck location. This
heart rate assessment, movement detection and GPS information. For concern was raised during a discussion with oil palm plantation man
example, operators’ heart rate can be monitored during operation in ager where thievery is one of the potential issues in the supply chain of
plantation and processing plants to ensure good health status and fresh fruit bunches. Currently, all fresh palm fruits are weighted before
wearable sweat sensors can be incorporated to detect the physical state departure from the plantation site, and the weight is measured once
of the wearer with health diagnostics [74,75]. Smartwatches were also again at processing mill upon arrival to confirm the exact sales amount.
developed to detect falls, thus allowing first aid response teams to Often the amount is not tally due to potential losses during trans
achieve a swift response when the farmer or user has been determined to portation or thievery. The incorporation of IoT weight and GPS sensors
have had a serious accident, especially during operation at height [76]. on delivery trucks may aid to minimise this problem. For example, if
Apart from that, wearable biosensors have been slowly introduced in the weight losses are detected when the truck is in motion, most likely the
healthcare industry to provide a unique personalised healthcare plan to losses may occur due to transportation issues such as sharp corners,
suit individual needs based on their health conditions [77]. Tracing the bumpy road conditions or speeding. On the other hand, if the weight
operators’ activity via biosensor in smartwatch will not only enhance losses are detected when the truck is in station mode or the truck loca
safety feature but can also be incorporated with insurance policy based tion deviates from the planned route based on GPS detection, an
on the specific activity performed by operators with a robust tracking investigation can be initiated for potential theft.
system and record. A similar type of application has been studied to improve traceability
With the regular monitoring and recording, application of smart by IoT devices. For instance, a radio-frequency identification system is
monitoring system can be extended to various aspects. For example, used in a container logistic system to identify different transport units
smart monitoring to improve product traceability which will be dis for monitoring and management purposes, improve efficiency, and
cussed in detail in Section 3.6. Besides, a smart contract between mul prevent container loss [82]. A study based on wood furniture supply
tiple stakeholders can be implemented, such as a dynamic incentive chain concluded that the investment cost for product tracking system
reward and healthcare coverage system based on farmers daily working should be equitably distributed among all stakeholders to improve the
activities tracked by wearable devices to identify scofflaws among their feasibility of implementation [83]. With such traceability capability
peers. This would minimise the issue of employees intentionally remove from Industry 4.0 technology, a more robust tracking system for CSPO
the smart wearable devices to temper with the tracking system as the can help to tackle the issue of palm biodiesel production in the European
failure of tracking their activities could affect the incentive and countries.
healthcare benefits. Further discussion on the smart contract application
will be presented in Section 3.7. 3.7. Smart contract-based implementation
3.6. Traceability-based implementation A smart contract is the concept of an autonomous system that is
designed to conduct a transaction without human intervention. It is
As discussed earlier, traceability of palm oil source is one of the key usually coded parameters written into the blockchain to perform specific
challenges to promote CSPO due to the limitation of the current system transaction if certain requirement or condition is met prior (contractual)
to differentiate sustainable and non-sustainable palm oil from the huge agreement from all parties [84]. Smart contract applications are also
amount of SMEs involved in the supply chain [8]. Besides, prolonged included in IoT, system security, and reduce the financial risk [85]. For
storage of palm fruit and oil will affect the oil extraction yield. These example, incorporation IoT and smart contract to enable automated
issues were addressed in the HAZOP Recommendations 9 (Blockchain transactions or dedicated payment schemes. In line with the Recom
technology to enhance the traceability of products across the supply mendation 4 (Health care devices such as smartwatches to monitor
chain) and 19 (Incorporate blockchain and IoT technology to ensure health status and location of worker) in Table 2, a more efficient and fair
each product is distributed within the maximum allowable storage payment system can be programmed based on farmers’ health condition
time). The recommendations suggested that the traceability of palm and activity pattern in plantation to replace the conventional daily fix
oil-based products needs to be improved. In order to prove the palm oil is rate scheme. Such implementation can be used to differentiate
sustainable, a robust tracking system that records the product lifecycle is committed employees from uncommitted employees and provide an
required to show the environmental impacts from each process. For adequate reward (or penalty), respectively. From Recommendation 11
proper recording and tracking purposes, blockchain has shown an (Smart contract to determine the premium price for good quality
8
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
products) and 16 (Incorporate smart contract to increase the selling as compared to centralised or multi-agent distributed architectures [95].
price of fresh palm fruit for SMEs with consistent supply), different A five-level trust model is proposed for cloud-edge based infrastructure
pricing of crude palm oil (CPO) based on its properties, quality, and to allow data owners to determine the trust level of each party. This
consistency of supply can be embedded into a smart contract to ensure a increases the flexibility of the model to react differently in cases which
consistent supply of premium palm oil, especially from SMEs. Farmers trust level can be adjusted based on specific action or performance [96].
that are able to produce premium palm oil that fulfils CSPO standard As a robust and high-end cybersecurity system can be very expensive, a
will be automatically awarded the premium price in the smart contract study proposed an attacker-defender model to evaluate the strategy for
system to encourage SMEs to initiate CSPO certification and continue defending assets in a digital logistic network [97]. Similarly, Game
with CSPO production. Theory has been applied to evaluate and optimise cloud security level
Similar to the traceability function discussed in Section 3.6, a robust and its pricing strategy [98].
smart contract system should be incorporated together with IoT-enabled From the discussion, numerous efforts have been made by re
devices to allow for automatic data monitoring and recording to avoid searchers to ensure safe implementation of Industry 4.0 technologies.
error from manual operation. For example, IoT-enabled containers are Simultaneous implementation of cybersecurity and other Industry 4.0
utilised to monitor shipment conditions including temperature, location, technologies in the palm oil industry should be performed i) to avoid the
pressure (for open/close detection of a pressurised container), and ac introduction of new forms of threat, and ii) to minimise operational risk
celeration for sudden fall or drop detection [86]. If any of the sensors in future. However, based on industry input, the processing information
detected violation of the agreed shipping condition, the information and knowledge in the palm oil industry are typically standard across
could be used to adjust the shipping cost as per the smart contract agreed different companies. Less effort and investment have been used to
by all stakeholders. This can be translated into the palm oil industry improve the security of information protection, except for the protection
where any unsustainable practice (e.g., deforestation which increases of research and development information due to potential patent gen
carbon release) is detected, the premium price of CSPO will, therefore, eration that may provide significant income to the company. Thus, any
not be awarded. Application of smart contract in a blockchain-based form of cybersecurity measure should be given priority to the protection
platform has shown promising results to improve the efficiency of of the research and development department during the transformation
instant decision making and support for a better quality of product and of Industry 4.0 in the palm oil industry.
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 4. Potential Industry 4.0 implementation in the palm oil
process stage. Payment system for cloud services with smart contract industry
incorporation was evaluated in a study. Two models were proposed,
including a proof-based verification model which only feasible if all HAZOP approach is generally used in the chemical engineering in
parties are assumed to be honest; and a replication-based verification dustry to review and improve process safety based on P&IDs. A new
model which works best when fines are imposed to cheating party [87]. version of P&ID is produced to incorporate practical recommendations
The concept of Game Theory is incorporated into a model to mitigate from the HAZOP review to reflect the changes for design improvement.
cheating and illegal behaviour to reduce the success rate of criminal By adopting the same philosophy, an updated block flow diagram (based
smart contract [88]. Another study discussed the application of smart on Fig. 2) is proposed to reflect the potential Industry 4.0 technologies
contract in an autonomous delivery system which includes consider implementation in the palm oil industry, as shown in Fig. 3. The pro
ation of supply chain performance, legal aspects and product custom posed potential implementations are based on the discussion of various
isation [89]. The smart contract also can be incorporated into financial Industry 4.0 technologies in Sections 3.1–3.8. A total of 13 Industry 4.0
loan management where this technology can be beneficial to handle features have been proposed. These features include general features
subsidies or loan among SMEs [90]. For example, the loan interest rate is that can be implemented across all stakeholders (Feature 1 to 3) and
imposed based on CSPO certification status to promote the initiation for specific features that can be adapted in particular process stages
CSPO certification where a lower interest rate will be imposed for those (Feature 4 to 13). These features are proposed to tackle the current
companies that are committed to fulfil the standard. challenges in the palm oil industry including traceability of CSPO,
inaccurate forecasting, poor labour and soil nutrient management, un
3.8. Cybersecurity-based implementation planned process downtime, safety and healthcare [26,29,45,81,99].
Feature 1 (Smart contract for fair pricing) in Fig. 3 suggests that
Despite the advantages of applying Industry 4.0 technologies to blockchain technology can be implemented across all stakeholders, from
improve process efficiency, the system may be exposed to higher secu upstream palm oil processes to downstream market to improve product
rity threat due to the interlinked nature of infrastructure. Cybersecurity lifecycle traceability and transaction security [78,93]. If blockchain
issues are related to several types of cyber-attacks, including viruses, technology is implemented across all the palm oil industry stakeholders,
Trojan horse, worms, phishing, Denial-of-Service attack, illegal access an industry-wide smart contract can be configured. An Industry 4.0
and control system attacks [91]. If any form of cybersecurity threat enabled industry can systematically generate fair transactions among
manages to infiltrate the system of IoT devices, the rest of the stakeholders based on their product sustainability performance, CSPO
inter-linked devices are prone to the risk of sabotage as well. This certification, the carbon footprint from processes and logistics, and so
problem has been addressed in the HAZOP review in Recommendation cial impacts. Premium price or subsidy will be awarded or removed
14 (Improve cybersecurity, such as blockchain for decentralised depending on the traceable parameters. The improved traceability
network). Numerous researchers have proposed various methods to function is critical to convince users, especially for palm biodiesel pro
evaluate and strengthen cybersecurity. A threat model is proposed to duction and usage in European countries, to show that the product
conduct a risk analysis of an IoT system to assess cyber threats and lifecycle is fulfilling the required sustainability standard. Feature 2
propose countermeasures [92]. Blockchain technology is incorporated (Traceability of product losses) proposes a product tracing system for
into the healthcare system to address the concern of information leakage logistics to avoid thievery, as discussed in Section 3.6. This will be
due to the decentralised nature of the blockchain system that resists to relatively easy to be implemented by a logistics company by installing
cyber hack [93]. Various approaches have been discussed to improve IoT-enabled sensors in delivery trucks or containers [82,83,89]. On the
security issues in cloud, edge and fog computing platforms [94]. other hand, the tedious digitalisation of printed documents creates
Blockchain technology can also be incorporated with smart contract reluctancy for SMEs to adopt Industry 4.0 revolution. As such, the re
system to enhance operational security in battery energy storage system view suggests using AI and deep learning technology in image process
from cyber-attack. The simulation result shows significant improvement ing to digitalise document as presented by Feature 3 (Digitalisation of
9
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
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 such as effectiveness in harvesting, duration of unplanned break time,
the palm oil industry as a supporting feature during the transition period awareness to safety, and health status. This feature is not recommended
toward Industry 4.0. to be implemented in the downstream processing plants because palm
Other features proposed from this study are more suitable to improve oil mill and refinery plant are usually operated in a more defined
specific palm oil processes, as shown on the right-hand side of Fig. 3. By boundary where supervision of workers is relatively easier as compared
incorporating IoT devices and cloud computing at oil palm plantations, to plantation sites.
the efficiency of plantation management can be improved. Feature 4 Other Industry 4.0 technologies can be incorporated into down
(Smart CPO production forecast) suggests that an enhanced forecast stream processes such as palm oil refineries to produce refined, bleached
system can be used for better yield prediction by installing IoT devices and deodorised palm oil for palm biodiesel, food, and chemicals pro
for local weather; and soil detections to induce smart farming for better duction. Feature 9 (Dynamic process optimisation) proposes a dynamic
water and fertiliser management, as suggested by Feature 5 (Smart process optimisation and control system to improve the efficiency of the
farming/irrigation) [27,36]. The data processing is proposed to be processes. This requires the installation of IoT sensors in the processes
conducted via cloud computing due to the consideration of relatively and computational facilities to analyse the dynamic fluctuation of big
low accessibility to powerful computing facilities at the plantation site. data. For example, in biodiesel production from palm oil and waste
Features 6 (Improve harvesting planning/forecast) and 7 (Optimum frying oil, the mixing ratio of the oils can be optimised based on the
harvesting route) are proposed to improve the efficiency of palm fruit dynamic changes of the feed properties and produced biodiesel prop
harvesting at the plantation site. The installation of IoT imaging devices erties [107]. Other alternative improvements include the optimisation
to detect the ripeness of palm fruit has been conducted by various re of the transesterification process in palm biodiesel production based on
searchers [49–51]. By combining the IoT devices at plantation sites for methanol to oil ratio, amount of catalyst, reaction temperature, and
weather forecast and palm fruit ripeness detection, a more accurate energy efficiency improvement [108]. With IoT sensors installed across
harvesting plan can be prepared to reduce labour cost as proposed in the processing system, two additional features can be configured as
Feature 6. A recent study suggested an inductive sensor system with up stated in Feature 10 (Equipment defect detection) and 11 (Remote ac
to 100% accuracy in detecting palm fruit ripeness and an accurate es cess of process control). Feature 10 proposes a defect equipment
timate of harvest time [101]. This feature can be further improved by detection system to enable smart maintenance management to minimise
incorporating AI and deep learning technology to optimise harvesting unplanned downtime [44,45]. A more advanced detection system may
route which aims to minimise the distance travelled by farmers within utilise image processing technologies (e.g., photography and CCTV) to
the plantation site as proposed in Feature 7 (Optimum harvesting route) determine any piping defect within the process plant, as discussed in
[102,103]. Optimisation of palm fruit harvesting route is similar to Section 3.3 [55]. Utilising the connection of IoT sensors with the control
handling cold supply chain that the transported material’s quality will system, remote access to the system can be configured as suggested in
deteriorate over time [104]. Feature 8 (Smart contract for price pri Feature 11. This may reduce the requirement of senior engineer at the
cing/salary) introduces wearable IoT devices with biosensors to monitor palm oil processing plant, especially during the night shift. As a result,
the health conditions and activities of farmers at plantation sites [75,76, the overall operation cost can be minimised; while enabling quick
105,106]. Such devices can be configured with smart contracts where remote respond during an emergency. Next, Features 12 (Smart restock
the workers are awarded and insured based on their daily performances notification) and 13 (Smart production management) are proposed to
10
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
install imaging processing and IoT devices to monitor the flow of consequently increase sustainable palm biodiesel production. Nonethe
palm-based products at the retailer market. This aims to create a smart less, investment into IoT sensors for detection and data collection is
marketing system where automated product replenishment can be necessary for the Industry 4.0 transformation. In order to analyse the
configured. In an ultimate scenario where the processing plants and massive amount of data from IoT devices, cloud computing was pro
markets are connected with IoT systems, smart production planning can posed for big data analysis instead of physical computing facilities due to
be created to avoid oversupply or undersupply of palm-based product almost none of the existing oil palm plantation and processing plant has
and storage requirement can be minimised. adequate computing capability for AI and big data analysis. The imaging
From the discussion, the majority of the proposed Industry 4.0 fea processing technology was proposed to improve the operational effi
tures require the integration of IoT devices. Due to the interconnected ciency, such as detection of fruit ripeness and faulty equipment.
nature of IoT system, the enhancement of cybersecurity of the palm oil In conclusion, this review paper has demonstrated the advantages of
industry should be given priority in the transformation. This should be HAZOP review approach in evaluating the status of the palm oil industry
highlighted to protect information breaches such as research effort, in the revolution of Industry 4.0. This article discussed the current
potential patent, personnel information (from wearable IoT devices), problems and the conventional mitigation measures in the palm oil in
CCTV footage, and process sabotage. dustry and explored the possible application of new Industry 4.0 tech
nologies to improve the system. Nonetheless, the review was limited to a
5. Conclusions macro-level of analysis for each stage of processes. A more detailed
consideration can be incorporated in future work. Instead of performing
This paper implemented a HAZOP approach to review the current a black box review of each process stage (e.g. the refinery), a more
state of the palm oil industry in Industry 4.0 revolution and develop detailed HAZOP literature review can be conducted to include various
ment. The state-of-the-art HAZOP review approach provides an in-depth process systems or equipment to provide a more holistic implementation
investigation of the system using a guidewords-based brainstorming of Industry 4.0 features in the palm oil industry. Palm waste such as
evaluation method. A total of 23 recommendations were proposed to palm biomass can also be included in a future study to understand the
highlight the potential gaps to incorporate Industry 4.0 technologies in overall footprint and the challenges toward palm oil industry 4.0.
the palm oil industry, including the use of IoT devices, smart image
processing technology, smart monitoring, smart forecasting, cloud CRediT authorship contribution statement
computing, traceability, smart contract, and cybersecurity. Analysis
from the review suggested that the combination of several Industry 4.0 Chun Hsion Lim: Conceptualization, Methodology, Investigation,
technologies is required. For example, the combination of IoT devices Resources, Writing - original draft, Writing - review & editing, Visuali
and cloud computing can enable smart CPO production forecast; while zation, Project administration, Funding acquisition. Steven Lim:
the combination of IoT devices, image processing and cloud computing Investigation, Resources, Funding acquisition. Bing Shen How: Inves
can optimise harvest route in oil palm plantation. This study had iden tigation, Resources, Writing - review & editing. Wendy Pei Qin Ng:
tified a total of 13 potential Industry 4.0 features to be implemented Investigation, Resources. Sue Lin Ngan: Investigation, Resources,
across all or specific palm oil industry stakeholders. A mapping diagram Writing - review & editing. Wei Dong Leong: Investigation, Resources.
was presented to show the possible improvement across the supply chain Hon Loong Lam: Investigation, Resources.
of palm-based products. Several proposed features are beneficial for the
industry as a countermeasure to the movement of banning palm oil Declaration of competing interest
usage in European countries due to the sustainability issue. Blockchain
technology from Industry 4.0 provides a secure transaction and data The authors declare that they have no known competing financial
recording to improve the traceability of the products. The imple interests or personal relationships that could have appeared to influence
mentation of such technology across palm oil industry stakeholders of the work reported in this paper.
fers a robust traceable system to develop sustainable profile and
certification of the product. This helps to promote CSPO production, and
Table A.1
HAZOP worksheet for Node 1
Node 1
Description: Palm oil upstream process including plantation, harvesting and milling
1. Feed 1. Fluctuation of 1. Uncertainty of weather 1. Harsh weather leading to 1. Regional weather prediction 1. Incorporation of localised
quantity affecting plantation yield low production yield. model based on satellite data, IoT devices to improve
and harvest activities. sea water level, sea buoys, etc. the accuracy of weather
prediction.
2. Non-optimum water/ 1. Too much or too little supply 1. Watering of plantation side 2. Incorporation of localised
fertiliser usage. of water/nutrient resulting based on historical rain fall IoT devices to detect soil
lower yield of oil. quantity. properties for moisture
2. Over supply of water/ 2. Fertiliser usage based on soil and nutrient.
fertiliser resulting higher and leaf analysis or
unnecessary cost. recommendation of fertiliser
supplier.
3. Scheduled annual fertiliser
application.
2. Inconsistency 1. Unhealthy palm tree. 1. Unhealthy palm tree leading 1. Satellite image to identify 3. AI imaging processing
of quality to potential low production unhealthy tree based on leaf technology to determine
yield and infection to other colour. tree health status locally.
trees.
2. Old tree. 1. Old tree causing low fruit 1. Scheduled replantation cycle
yield, resulting in non- to ensure optimum yield.
(continued on next page)
11
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
Description: Palm oil upstream process including plantation, harvesting and milling
12
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
Description: Palm oil upstream process including plantation, harvesting and milling
harvested palm oil fruit spec/contamination of palm minimise chances of off-spec price for good quality
bunches (for example due to oil. product. products.
weather condition). 2. Unfair costing (same selling
price across different quality
of fruit bunches).
3. Storage 1. Quality degradation of palm 1. Long storage promotes 1. Pretreatment such as
oil fruit bunches due to bacteria growth, resulting sterilization in milling process
prolong storage. off-spec product. to minimise impact of
microorganism.
4. Product 1. Unable to deliver product on 1. Potential higher 1. Location of plantation side
distribution time. transportation due to the and mill is relatively close.
delayed deliver (such as 2. Use of GPS to detect road
traffic, vehicle breakdown). condition.
2. Prolonged delivery time 3. Transport during off-peak
reduces quality of fresh palm hours to avoid heavy traffic.
fruit (increased of FFA and
bacteria) leading to
potential off-spec product.
2. Unable to trace fruit losses 1. Loss of fresh palm fruit 1. Truck loading weight check 12. IoT devices and GPS
during transportation during transportation (due prior departing from system to detect live
to insecure loading or plantation and upon receival loading weight and
thievery) leading to profit at mill to determine losses truck location during
losses. during transportation. distribution.
4. Cost 1. Fluctuation of 1. Higher cost of organic 1. Potential high risk of 1. Partial testing of new fertiliser
feed cost fertiliser compared to investment into new type of to compare and monitor the
inorganic fertiliser. fertiliser without justified improvement, such as oil yield
return. to minimise investment risk.
2. Fluctuation of 1. Dynamic price of CPO. 1. Potential changes of palm 1. Forecasting of palm oil Previous Recommendations
product price fruit bunches price leading production to estimate CPO 8 and 11 are applicable here:
to unsecure or unfair pricing price in future for planning 8. IoT and big data analysis
from dealers. purpose. to improve forecasting
2. Unable to differentiate techniques for fresh
prices for different quality of palm fruit generation.
product 11. Smart contract to
determine the premium
price for good quality
products.
3. Processing cost 1. Unplanned shutdown of 1. Emergency shutdown or less 1. Current practice to ensure
process. feed scenario leading to adequate amount of EFB to be
higher production cost to stored to maintain boiler
maintain utilities supply, operation for specific duration
especially boilers. However, as per company operation
the effect is less significant philosophy.
as majority of palm oil mill is
using EFB as boiler feed.
5. Accuracy of 1. Traceability of 1. Manual documentation and 1. Potential loss of tracing of 13. Provide tools for
information resources recording method. each batch of product digitalisation of printed
resulting wrong documents.
documentation and Previous Recommendation 9
recording. is applicable here:
9. Blockchain technology
to enhance traceability
of products across the
supply chain.
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 1. Potential significant losses in 1. Private and confidential 14. Improve cybersecurity,
development information. R&D, such as patent. contract with employees. such as blockchain for
2. Encrypted computer/laptop to decentralised network.
avoid information extraction
or sharing prior approval.
13
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
Table A.2
HAZOP worksheet for Node 2
Node 2
Description: Processes of palm oil including refinery, bio-diesel production, and food/pharmaceutical production
1. Feed 1. Fluctuation of 1. Impact of regional CPO 1. Major impact to process, 1. Storage system of CPO 15. Improvement of CPO
quantity production yield such as potential process shutdown. for consistent supply to production forecast for better
harsh weather, diseases, refinery and supply chain and storage
seasonality yield etc. downstream processes. management.
2. Inconsistent supply from 1. Potential shutdown of milling 1. Storage system of CPO 16. Incorporate smart contract to
various suppliers, process during low resource of for consistent supply to increase the selling price of
especially SMEs. fruit bunches resulting higher refinery and fresh palm fruit for SMEs with
cost for starting-up and shut downstream processes. consistent supply.
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.
2. Inconsistency 1. Different physical and 1. Potential product off-spec,
of quality chemical properties of however this is very unlikely
CPO from different due to establishment of
suppliers. adequate pretreatment
processes.
2. Different sustainability 1. Potential unable to achieve 1. Palm oil source is Previous Recommendations 9 and
profile from different MSPO requirement if the limited suppliers which 10 are applicable here:
suppliers due to source has high environmental have MSPO 9. Blockchain technology to
different farming, impact. certification. enhance traceability of products
harvesting practices 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.
3. Storage 1. Prolonged storage of 1. Potential off-spec of CPO 1. Monitoring and control
CPO prior processing to affecting efficiency of down of storage condition.
downstream product. stream processes, however this 2. Various studies for
is considered as unlikely event optimum palm oil
due to CPO is relatively stable storage has been
for storage. conducted.
4. Various 1. Involvement of multiple 1. Potential inconsistency of raw
suppliers suppliers, especially material quality, which has
from SMEs. same consequences as
“Inconsistency in quality” in
Node 2.
2. Processing 1. Monitoring of 1. Visual inspection and 1. Localise inspection which may 1. Provide adequate 17. Incorporation of IoT devices in
efficiency process monitoring of process cause potential exposure of protection such as the process to enable remote
condition. hazardous environment to personnel protective monitoring and access for
personnel. equipment when process control.
2. Visual inspection may not able conduct routine Previous Recommendation 7 is
to detect process deviation monitoring at applicable here:
accurately. 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.
2. Process 1. Localised process 1. Potential cascade effect from 1. Mitigation control Previous Recommendation 17 is
intervention/ control for each control single control failure which system for cascade applicable here:
control loop. may affects multiple process effect, such as 17. Incorporation of IoT devices in
equipment and possible sequential unit the process to enable remote
process shutdown. shutdown due to monitoring and access for
process deviation. process control.
2. No communication 1. Unable to configure preventive 1. Current process control 18. Incorporation of big data
between process control control between independent philosophy is based on analysis, cloud computing and
loops. control system/equipment. 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.
3. Maintenance 1. Unplanned process 1. Frequent unplanned shutdown Previous Recommendation 7 is
shutdown due to cause loss of production and applicable here:
equipment breakdown. potential equipment damage. 7. Incorporate IoT and big data
analysis on machinery, detectors,
controllers to determine
(continued on next page)
14
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
Description: Processes of palm oil including refinery, bio-diesel production, and food/pharmaceutical production
15
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
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
Description: Market for products/end users for palm oil product including palm bio-diesel, edible palm oil and palm chemicals products
17
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
[29] Carron MP, Pierrat M, Snoeck D, Villenave C, Ribeyre F, Surhadi, et al. Temporal (HSI) approach. IOP Conf Ser Earth Environ Sci 2016;37:012039. https://doi.org/
variability in soil quality after organic residue application in mature oil palm 10.1088/1755-1315/37/1/012039.
plantations. Soil Res 2015. https://doi.org/10.1071/SR14249 Temporal. [54] Liu J, Wang X, Wang T. Classification of tree species and stock volume estimation
[30] Rhebergen T, Zingore S, Giller KE, Frimpong CA, Acheampong K, Ohipeni FT, in ground forest images using Deep Learning. Comput Electron Agric 2019;166:
et al. Closing yield gaps in oil palm production systems in Ghana through Best 105012. https://doi.org/10.1016/j.compag.2019.105012.
Management Practices. Eur J Agron 2020;115:126011. https://doi.org/10.1016/ [55] Yin X, Chen Y, Bouferguene A, Zaman H, Al-hussein M, Kurach L. A deep learning-
j.eja.2020.126011. based framework for an automated defect detection system for sewer pipes.
[31] Rhebergen T, Fairhurst T, Giller KE, Zingore S. The influence of water and Autom ConStruct 2020;109:102967. https://doi.org/10.1016/j.
nutrient management on oil palm yield trends on a large-scale plantation in autcon.2019.102967.
Ghana. Agric Water Manag 2019;221:377–87. https://doi.org/10.1016/j. [56] Nturambirwe JFI, Opara UL. Machine learning applications to non-destructive
agwat.2019.05.003. defect detection in horticultural products. Biosyst Eng 2019;189:60–83. https://
[32] Comte I, Colin F, Grünberger O, Follain S, Whalen JK, Caliman JP. Landscape- doi.org/10.1016/j.biosystemseng.2019.11.011.
scale assessment of soil response to long-term organic and mineral fertilizer [57] Kim DH, Lee S, Jeon J, Song BC. Real-time purchase behavior recognition system
application in an industrial oil palm plantation, Indonesia. Agric Ecosyst Environ based on deep learning-based object detection and tracking for an unmanned
2013;169:58–68. https://doi.org/10.1016/j.agee.2013.02.010. product cabinet. Expert Syst Appl 2020;143:113063. https://doi.org/10.1016/j.
[33] Pauli N, Donough C, Oberthur T, Cock J, Verdooren R, Rahmadsyah, et al. eswa.2019.113063.
Changes in soil quality indicators under oil palm plantations following [58] Cilia ND, Stefano C De, Fontanella F, Marrocco C, Molinara M, Freca AS Di. An
application of “best management practices” in a four-year field trial. Agric end-to-end deep learning system for medieval writer identification. Pattern
Ecosyst Environ 2014;195:98–111. https://doi.org/10.1016/j.agee.2014.05.005. Recogn Lett 2020;129:137–43. https://doi.org/10.1016/j.patrec.2019.11.025.
[34] Boafo DK, Kraisornpornson B, Panphon S, Owusu BE, Amaniampong PN. Effect of [59] Dasgupta J, Bhattacharya K, Chanda B. A holistic approach for Off-line
organic soil amendments on soil quality in oil palm production Daniel. Elsevier; handwritten cursive word recognition using directional feature based on Arnold
2019. https://doi.org/10.1016/j.apsoil.2019.09.008. transform. Pattern Recogn Lett 2016;79:73–9. https://doi.org/10.1016/j.
[35] Glaroudis D, Iossifides A, Chatzimisios P. Survey, comparison and research patrec.2016.05.017.
challenges of IoT application protocols for smart farming. Comput Network 2020; [60] Zhang X, Bengio Y, Liu C. Online and offline handwritten Chinese character
168:107037. https://doi.org/10.1016/j.comnet.2019.107037. recognition: a comprehensive study and new benchmark. Pattern Recogn 2017;
[36] Krishnan RS, Julie EG, Robinson YH, Raja S, Kumar R, Huy P, et al. Fuzzy logic 61:348–60. https://doi.org/10.1016/j.patcog.2016.08.005.
based smart irrigation system using internet of things. J Clean Prod 2020;252: [61] Ali O, Osmanaj V. The role of government regulations in the adoption of cloud
119902. https://doi.org/10.1016/j.jclepro.2019.119902. computing: a case study of local government. Comput Law Secur Rev 2020;36:
[37] Khachatryan H, Suh DH, Xu W, Useche P, Dukes MD. Towards sustainable water 105396. https://doi.org/10.1016/j.clsr.2020.105396.
management: preferences and willingness to pay for smart landscape irrigation [62] Sun P. Security and privacy protection in cloud computing: discussions and
technologies. Land Use Pol 2019;85:33–41. https://doi.org/10.1016/j. challenges. J Netw Comput Appl 2020;160:102642. https://doi.org/10.1016/j.
landusepol.2019.03.014. jnca.2020.102642.
[38] Shi J, Wu X, Wang X, Zhang M, Han L, Zhang W, et al. Determining threshold [63] Khayer A, Talukder S, Bao Y, Hossain N. Cloud computing adoption and its
values for root-soil water weighted plant water de fi cit index based smart impact on SMEs’ performance for cloud supported operations: a dual-stage
irrigation. Agric Water Manag 2020;230:105979. https://doi.org/10.1016/j. analytical approach. Technol Soc 2020;60:101225. https://doi.org/10.1016/j.
agwat.2019.105979. techsoc.2019.101225.
[39] Wang J, Chen M, Zhou J, Li P. Data communication mechanism for greenhouse [64] Shaharum NSN, Shz M, Ghani WAWAK, Samsatli S, Al-Habshi MMA, Yusuf B. Oil
environment monitoring and control: an agent-based IoT system. China palm mapping over Peninsular Malaysia using Google Earth Engine and machine
Agricultural University; 2019. https://doi.org/10.1016/j.inpa.2019.11.002. learning algorithms. Remote Sens Appl Soc Environ 2020;17:100287. https://doi.
[40] Lee I. The Internet of Things for enterprises: an ecosystem, architecture, and IoT org/10.1016/j.rsase.2020.100287.
service business model. Internet Things 2019;7:100078. https://doi.org/ [65] Lee JSH, Wich S, Widayati A, Koh LP. Detecting industrial oil palm plantations on
10.1016/j.iot.2019.100078. Landsat images with Google Earth Engine. Remote Sens Appl Soc Environ 2016;4:
[41] Zou Y, Yu L, Tso GKF, He K. Risk forecasting in the crude oil market: a multiscale 219–24. https://doi.org/10.1016/j.rsase.2016.11.003.
Convolutional Neural Network approach. Physica A 2020;541:123360. https:// [66] Tehlah N, Kaewpradit P, Mujtaba IM. Artificial neural network based modeling
doi.org/10.1016/j.physa.2019.123360. and optimization of refined palm oil process. Neurocomputing 2016;216:
[42] Ahmed K, Sachindra DA, Shahid S, Iqbal Z, Nawaz N. Multi-model ensemble 489–501. https://doi.org/10.1016/j.neucom.2016.07.050.
predictions of precipitation and temperature using machine learning algorithms. [67] Wanrosli WD, Zainuddin Z, Ong P, Rohaizu R. Optimization of cellulose
Atmos Res 2020;236:104806. https://doi.org/10.1016/j.atmosres.2019.104806. phosphate synthesis from oil palm lignocellulosics using wavelet neural networks.
[43] Chapman R, Cook S, Donough C, Lim YL, Ho PVV, Lo KW, et al. Using Bayesian Ind Crop Prod 2013;50:611–7. https://doi.org/10.1016/j.indcrop.2013.08.048.
networks to predict future yield functions with data from commercial oil palm [68] Ahmad AL, Azid IA, Yusof AR, Seetharamu KN. Emission control in palm oil mills
plantations: a proof of concept analysis. Comput Electron Agric 2018;151: using artificial neural network and genetic algorithm. Comput Chem Eng 2004;
338–48. https://doi.org/10.1016/j.compag.2018.06.006. 28:2709–15. https://doi.org/10.1016/j.compchemeng.2004.07.034.
[44] Kobbacy KAH. Application of artificial intelligence in maintenance modelling and [69] Kesarwani A, Khilar PM. Development of trust based access control models using
management. IFAC Proc Vol 2012;4:22–3. https://doi.org/10.3182/20121122-2- fuzzy logic in cloud computing. King Saud University; 2019. https://doi.org/
ES-4026.00046. IFAC; 2012, pp. 54–9. 10.1016/j.jksuci.2019.11.001.
[45] Lee WJ, Wu H, Yun H, Kim H, Jun MBG, Sutherland JW. Predictive maintenance [70] Novais L, Maqueira JM, Ortiz-Bas Á. A systematic literature review of cloud
of machine tool systems using artificial intelligence techniques applied to computing use in supply chain integration. Comput Ind Eng 2019;129:296–314.
machine condition data. Procedia CIRP 2018;80:506–11. https://doi.org/ https://doi.org/10.1016/j.cie.2019.01.056.
10.1016/j.procir.2018.12.019. [71] Ujjwal KC, Garg S, Hilton J, Aryal J, Forbes-smith N. Cloud Computing in natural
[46] Zhang X, Chen MY, Wang MG, Ge YE, Stanley HE. A novel hybrid approach to hazard modeling systems: current research trends and future directions. Int J
Baltic Dry Index forecasting based on a combined dynamic fluctuation network Disaster Risk Reduct 2019;38:101188. https://doi.org/10.1016/j.
and artificial intelligence method. Appl Math Comput 2019;361:499–516. ijdrr.2019.101188.
https://doi.org/10.1016/j.amc.2019.05.043. [72] Kumar M, Sharma SC, Goel A, Singh SP. A comprehensive survey for scheduling
[47] Villa-Henriksen A, Edwards GTC, Pesonen LA, Green O, Sørensen CAG. Internet of techniques in cloud computing. J Netw Comput Appl 2019;143:1–33. https://doi.
Things in arable farming: implementation, applications, challenges and potential. org/10.1016/j.jnca.2019.06.006.
Biosyst Eng 2020;191:60–84. https://doi.org/10.1016/j. [73] Parikh S, Dave D, Patel R, Doshi N. Security and privacy issues in cloud, fog and
biosystemseng.2019.12.013. edge computing. Procedia Comput Sci 2019;160:734–9. https://doi.org/
[48] Hormaza P, Fuquen EM, Romero HM. Phenology of the oil palm interspecific 10.1016/j.procs.2019.11.018.
hybrid Elaeis oleifera × Elaeis guineensis. Sci Agric 2012;69:275–80. [74] Jin X, Gu F, Niu J, Yu S, Ouyang Z. HRCal: an effective calibration system for
[49] Juman MA, Wong YW, Rajkumar RK, Goh LJ. A novel tree trunk detection heart rate detection during exercising. J Netw Comput Appl 2019;136:1–10.
method for oil-palm plantation navigation Mohammed. Comput Electron Agric https://doi.org/10.1016/j.jnca.2019.03.007.
2016;128:172–80. https://doi.org/10.1016/j.compag.2016.09.002. [75] Legner C, Kalwa U, Patel V, Chesmore A, Pandey S. Sweat sensing in the smart
[50] Rincón SM, Hormaza PA, Moreno LP, Prada F, Portillo DJ, García JA, et al. Use of wearables era: towards integrative, multifunctional and body-compliant
phenological stages of the fruits and physicochemical characteristics of the oil to perspiration analysis Christopher. Sensors Actuators A Phys 2019;296:200–21.
determine the optimal harvest time of oil palm interspecific OxG hybrid fruits. https://doi.org/10.1016/j.sna.2019.07.020.
Ind Crop Prod 2013;49:204–10. https://doi.org/10.1016/j.indcrop.2013.04.035. [76] Chen L, Li R, Zhang H, Tian L, Chen N. Intelligent fall detection method based on
[51] Makky M. A portable low-cost non-destructive ripeness inspection for oil palm accelerometer data from a wrist-worn smart watch. Measurement 2019;140:
FFB. Agric Agric Sci Procedia 2016;9:230–40. https://doi.org/10.1016/j. 215–26. https://doi.org/10.1016/j.measurement.2019.03.079.
aaspro.2016.02.139. [77] Papa A, Mital M, Pisano P, Giudice M Del. E-health and wellbeing monitoring
[52] Silalahi DD, Reano CE, Lansigan FP, Panopio RG, Bantayan NC. Using genetic using smart healthcare devices: an empirical investigation. Technol Forecast Soc
algorithm neural network on near infrared spectral data for ripeness grading of oil Change 2020;153:119226. https://doi.org/10.1016/j.techfore.2018.02.018.
palm (Elaeis guineensis Jacq.) fresh fruit. Inf Process Agric 2016;3:252–61. [78] Liu XL, Wang WM, Guo H, Barenji AV, Li Z, Huang GQ. Industrial blockchain
https://doi.org/10.1016/j.inpa.2016.10.001. based framework for product lifecycle management in industry 4.0. Robot
[53] Shabdin MK, Shariff ARM, Johari MNA, Saat NK, Abbas Z. A study on the oil palm Comput Integrated Manuf 2020;63:101897. https://doi.org/10.1016/j.
fresh fruit bunch (FFB) ripeness detection by using Hue, Saturation and Intensity rcim.2019.101897.
18
C.H. Lim et al. Renewable and Sustainable Energy Reviews 135 (2021) 110223
[79] Drljevic N, Aranda DA, Stantchev V. Perspectives on risks and standards that [95] Mhaisen N, Fetais N, Massoud A. Secure smart contract-enabled control of battery
affect the requirements engineering of blockchain technology. Comput Stand energy storage systems against cyber-attacks. Alexandria Eng J 2019;58:
Interfac 2020;69:103409. https://doi.org/10.1016/j.csi.2019.103409. 1291–300. https://doi.org/10.1016/j.aej.2019.11.001.
[80] Wang Z, Wang T, Hu H, Gong J, Ren X, Xiao Q. Blockchain-based framework for [96] Chadwick DW, Fan W, Costantino G, Lemos R De, Di F, Herwono I, et al. A cloud-
improving supply chain traceability and information sharing in precast edge based data security architecture for sharing and analysing cyber threat
construction. Autom ConStruct 2020;111:103063. https://doi.org/10.1016/j. information. Future Generat Comput Syst 2020;102:710–22. https://doi.org/
autcon.2019.103063. 10.1016/j.future.2019.06.026.
[81] Lomotey RK, Pry J, Sriramoju S. Wearable IoT data stream traceability in a [97] Cheung K, Bell MGH. Attacker – defender model against quantal response
distributed health information system. Pervasive Mob Comput 2017;40:692–707. adversaries for cyber security in logistics management : an introductory study.
https://doi.org/10.1016/j.pmcj.2017.06.020. Elsevier B.V.; 2019. https://doi.org/10.1016/j.ejor.2019.10.019.
[82] Madlenák R, Madlenáková L, Kolarovszká Z. System of management and [98] Kakkad V, Shah H, Patel R, Doshi N. A comparative study of applications of game
traceability of logisticitems through new technologies. Procedia - Soc Behav Sci theory in cyber A comparative study of applications of game theory in cyber
2016;230:128–35. https://doi.org/10.1016/j.sbspro.2016.09.016. security and cloud computing. Security and cloud computing. Procedia Comput
[83] Appelhanz S, Osburg V, Toporowski W, Schumann M. Traceability system for Sci 2019;155:680–5. https://doi.org/10.1016/j.procs.2019.08.097.
capturing, processing and providing consumer-relevant information about wood [99] Leegwater M, Duijn G van. 24 - traceability of RSPO-certified sustainable palm
products: system solution and its economic feasibility. J Clean Prod 2016;110: oil. Palm Oil Prod. Process. Charact. Uses 2012:713–36.
132–48. https://doi.org/10.1016/j.jclepro.2015.02.034. [100] Greenbaum NR, Jernite Y, Halpern Y, Calder S, Nathanson LA, Sontag DA, et al.
[84] Guadamuz A. All watched over by machines of loving grace: a critical look at Improving documentation of presenting problems in the emergency department
smart contracts. Comput Law Secur Rev 2019;35:105338. https://doi.org/ using a domain-specific ontology and machine learning-driven user interfaces. Int
10.1016/j.clsr.2019.105338. J Med Inf 2019;132:103981. https://doi.org/10.1016/j.ijmedinf.2019.103981.
[85] Zheng Z, Xie S, Dai H, Chen W, Chen X, Weng J, et al. An overview on smart [101] Sinambela R, Mandang T, Subrata IDM, Hermawan W. Application of an
contracts: challenges , advances and platforms. Future Generat Comput Syst inductive sensor system for identifying ripeness and forecasting harvest time of oil
2020;105:475–91. https://doi.org/10.1016/j.future.2019.12.019. palm. Sci Hortic (Amst) 2020;265:109231. https://doi.org/10.1016/j.
[86] Hasan H, Alhadhrami E, Aldhaheri A, Salah K, Jayaraman R. Smart contract- scienta.2020.109231.
based approach for efficient shipment management. Comput Ind Eng 2019;136: [102] He L, Xue M, Gu B. Internet-of-Things enabled supply chain planning and
149–59. https://doi.org/10.1016/j.cie.2019.07.022. coordination with big data services: certain theoretic implications. China Science
[87] Dorsala MR, Sastry VN, Chapram S. Fair payments for verifiable cloud services Publishing & Media Ltd.; 2020. https://doi.org/10.1016/j.jmse.2020.03.002.
using smart contracts. Comput Secur 2020;90:101712. https://doi.org/10.1016/ [103] Lioutas ED, Charatsari C. Land Use Policy Smart farming and short food supply
j.cose.2019.101712. chains: are they compatible? Land Use Pol 2020;94:104541. https://doi.org/
[88] Zhang L, Wang Y, Li F, Hu Y, Au MH. A game-theoretic method based on Q- 10.1016/j.landusepol.2020.104541.
learning to invalidate criminal smart contracts. Inf Sci 2019;498:144–53. https:// [104] Theeb N Al, Smadi HJ, Al-Hawari TH, Aljarrah MH. Optimization of vehicle
doi.org/10.1016/j.ins.2019.05.061. routing with inventory allocation problems in Cold Supply Chain Logistics.
[89] Prause G. Smart Contracts for Smart Supply Chains. IFAC Pap n.d.;52:2501–2506. Comput Ind Eng 2020;142:106341. https://doi.org/10.1016/j.cie.2020.106341.
https://doi.org/10.1016/j.ifacol.2019.11.582. [105] Abouzahra M, Ghasemaghaei M. The antecedents and results of seniors’ use of
[90] Wang H, Guo C, Cheng S. LoC - a new financial loan management system based on activity tracking wearable devices. Elsevier Ltd; 2019. https://doi.org/10.1016/j.
smart contracts. Future Generat Comput Syst 2019;100:648–55. https://doi.org/ hlpt.2019.11.002.
10.1016/j.future.2019.05.040. [106] Cosoli G, Spinsante S, Scalise L. Wrist-worn and chest-strap wearable devices:
[91] Srinivas J, Das AK, Kumar N. Government regulations in cyber security: systematic review on accuracy and metrological characteristics. Measurement
framework, standards and recommendations. Future Generat Comput Syst 2019; 2020;159:107789. https://doi.org/10.1016/j.measurement.2020.107789.
92:178–88. https://doi.org/10.1016/j.future.2018.09.063. [107] Almeida VF de, García-Moreno PJ, Guadix A, Guadix EM. Biodiesel production
[92] Casola V, Benedictis A De, Rak M, Villano U. Toward the automation of threat from mixtures of waste fish oil, palm oil and waste frying oil: optimization of fuel
modeling and risk assessment in IoT systems. Internet Things J 2019;7:100056. properties. Fuel Process Technol 2015;133:152–60. https://doi.org/10.1016/j.
https://doi.org/10.1016/j.iot.2019.100056. fuproc.2015.01.041.
[93] Tripathi G, Ahad MA, Paiva S. S2HS- A blockchain based approach for smart [108] Khoobbakht G, Kheiralipour K, Rasouli H, Rafiee M, Hadipour M, Karimi M.
healthcare system. Elsevier Inc.; 2019. https://doi.org/10.1016/j. Experimental exergy analysis of transesterification in biodiesel production.
hjdsi.2019.100391. Energy 2020;196:117092. https://doi.org/10.1016/j.energy.2020.117092.
[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