You are on page 1of 10

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/337220996

Developing AI Based Scheme for Project Planning by Expert Merging Revit and
Primavera Software

Conference Paper · March 2019


DOI: 10.1109/SSD.2019.8893274

CITATIONS READS

3 223

2 authors, including:

Karim Mohammed Aljebory


Al-Qalam University college
15 PUBLICATIONS   87 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Karim Mohammed Aljebory on 16 November 2022.

The user has requested enhancement of the downloaded file.


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

Developing AI Based Scheme for Project Planning by


Expert Merging Revit and Primavera Software

Karim M. Aljebory (1) Mustafa QaisIssam(2)


(1) (2)
Al Qalam University College, kmjebory@gmail.co m mustafa.qaisisam@yahoo.com

product quality. Planning, currently, depends and


Abstract: Project control and monitoring tools
relies upon manual formulation, tracking of plans
are based on expert judgment and parametric
and is usually performed in an intuitive and
tools. Some of these tools are Artificial
unstructured fashion [1, 8]. Formulation of a plan
Intelligence techniques. Today’s project
implies identifying work tasks, generating
planning tools require the generation and
activities list, predicting relationships between
sequencing of each activity.
these activities, determining required resources
Establishing interactive computer based and calculating partial budget to assist in
scheduled plan for a project requires not only expecting final project budget etc..., by a group of
injecting the information technology, but all engineers which called Planning Project Team
project activities and inserting their durations (PPT). The PPT almost carry out their job using
and resource requirements in organized and critical path method (CPM) to achieve good and
sequenced manner. New parametric Building successful plan. CPM helps the PPT in identifying
Information Modeling (BIM) software like which activities are critical and need to be in time
Revit can dramatically enhance planning according to the schedule.
productivity by substantially reducing the work This paper will focus on those using Artificial
necessary for preparing construction schedule. Intelligence algorithms and conducting project
The new parametric BIM can automate success estimation and concerns with:
exchange of information during design, - supporting automation of the construction
construction and service life of projects. process at its early planning stages.
In this paper, the Revit is connected to - achieving more realistic schedule for a
Primavera software by conducting expert construction project by exploiting a rich-data
system that supports automatic generation of model of BIM.
the project plan scheduling scheme of project - building a frame work of knowledge base
control functions. This will eliminate some of system that make use of BIM (Revit) data files and
the most tedious and time consuming steps in get access/ control to Primavera database.
creating a construction project schedule.
II. PLANNING AND SCHEDULING (PS)
Keywords- BIM, AI, Project Planning, Project
Control. Before starting any construction project, both
small and big, a guideline has to be developed.
The project plan passes through two main phases;
I. INTRODUCTION planning phase and scheduling phase. The
planning, consists of two main processes; choice
of construction technologies, and definition of
Planning in construction project is an essential, work tasks, while scheduling involves four main
crucial, fundamental and challenging activity in processes; sequencing of tasks, estimation of the
the management and execution of projects. A good required resources and durations for individual
plan is that safely execute the project in time and tasks, identification of the interactions or
budget constraints, and not compromising the constraints among the related tasks. The general

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 404


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

description for Planning and Scheduling processes simple division of the quantity by productivity, or
is described as follows. by referring to historical records of past projects.
An imperial method is to find the duration from
1- Define Construction Methods: selection of
expertise who are out in the field doing the work.
construction method to be replanned and decided.
It, directly, affects the cost estimate as well as the 6- Schedule development/ schedule generation:
schedule of the project. As an example, the Once the project has been organized into a
method of pouring concrete can be either by a sequence of activities and sub-activities, the PPT
pump or it can be transported in buckets. This needs to identify a logical order in which they
decision may, directly, affects the time duration should be sequenced, determine its resources,
and the budget needed for the completion of the duration, budget and in the shortest possible time.
activity. This is usually done by the critical path method
(CPM), network diagram, schedule compression,
2- Define the tasks schedule: decomposing the
resource levelling. With advances in technology,
project into activities and sub-activities. It is the
software packages provide better, faster and more
most cumbersome job to do, but this is the
accurate schedules.
important step assisting in the development of a
formal schedule of the project, because there can
be thousands of activities in a project. Schedule 7- Project Control: This involves monitoring the
tasks are usually organized in a project schedule project execution and insures that activities are at
using Work breakdown structure (WBS) which their scheduled time and within scheduled budget
organizes the sub-activities into levels that are to keep the project on schedule and within budget.
developed into a summary. Project breakdown
The developed system concerns only the core
structure is a tree form reflecting the work
factors in the first six processes to develop a
products on all levels produced by a project. While
realistic plan. The developed application extracts
the WBS defines activities which are required to
information from the BIM software as data-rich
enhance the work product to the required quality.
design elements and provides actual activities and
their sequencing for scheduling projects on
3- Sequence schedule tasks: Proper sequencing Primavera software package.
of the activities plays an important role in the
success or failure of the schedule. Activity
Facilities of Building Information Modelling:
sequence is the order in which each activity should
BIMis developed to simulate the construction and
be performed on the site. There are many factors
operation of projects. The BIM model is data-rich,
govern the sequencing and maintain structural
object-oriented, intelligent and parametric digital
integrity of activity, trade interaction, path
representation of the facility, from which views
interference and safety regulations.
and data appropriate to various users can be
extracted. The generated information can be used
4- Estimate resource requirement of each task: to make decisions and improve the process of
this involves deciding the type, number of delivering the facility. It is essential source of
workers, material and equipment. In this stage the information that is required for analysis, planning,
required resources to complete the project, and etc... [3, 6].
resources for each activity are identified. By
identifying the resources of the project,
bottlenecks related to resource allocation and Facilities of Primavera P6; this software has
activity sequencing are avoided. All these are done many facilities for project and time management
before calculating the duration of the activities. that meets various needs of project managers in
project plan designing and controlling. Although
Primavera is essentially designed for project-based
5- Estimate duration of each task: the duration organizations, it can be used for independent
of activity is the required time for completion. projects. The project elementary structure can be
Duration can be computed if the quantity of work fully defined by this software.
and the productivity of the crew are known by
Primavera P6 is designed for timing, controlling,

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 405


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

and estimating costs of projects. Some of the Artificial Intelligence (AI) in order to process data
objectives that are met using this software are: like human intelligence. Knowledge based expert
system (KBES) is defined as; AI computer
- Making strategic decisions
applications to solve complex problems in a
- Precise control of time up to the end of project particular domain, at the level of human
- Identifying the required resources, and resource intelligence and expertise. To this end, the
prioritization and estimation in long period of time research is exploring KBES as an approach to
extend the range of computer based applications in
- Reorganizing projects based on the new priority the construction project planning.
order without any negative influence on the
quality KBES technique applications and usefulness are
presented in a wide range of research studies.
- Reporting, auditing while project progressing Some applications were introduced;
- In-time manifesting the critical situations. CONSTRUCTION PLANEX developed by
Hendrickson (1987) as a frame based hierarchical
system for scheduling construction projects. It has
Facilities of Knowledge based Expert System
tables for functions like construction technology
(KBES): The KBES is proposed to perform all
selection, form project activities, duration
basic tasks of Human Planner/ Scheduler. A KBES
estimation, cost, predecessor selection, etc. [13].
can provide human expertise by; both the
knowledge engineering language and the program- Navinchandra, et al. (1988) developed aprogram
supporting environment. The basic structure of called Generation of Hierarchical networks for
KBES is as shown in Figure 1. construction (GHOST). The methodology
modifies the topology of construction activity
The AI application of KBES requires development
networks [10]. Knowledge is not used to build the
of a knowledge base that permits engineers to
network but to criticize it. It starts by assuming
interact with: the characteristic data, the
that all the activities will be executed
theoretical or simulation results, and the specific
simultaneously with a general network and
hypothesis of control measures. KBES have very
subdivides it into sub activities then modifies the
high potential for tackling problems that lack
network by examining what cannot be done in
explicit algorithms (e.g., problems for which a
parallel. This planning technique inputs a list of
numerical model does not exists). It retrieves data
activities and produces a schedule as output by
from many types of documents or project design in
setting up the predecessors among the activities.
BIM format, Geographical Information System
(GIS). The system involves building consistent Kartam and Levitt (1990) studied intelligent
components to maintain interaction between these planning of construction projects (SIPE), and
components and provide quick access to each proposed a new sequencing based on common
other. principles or constraints such as “supported by”,
“enclosed in” etc. SIPE can generate logically
correct activity networks from a description of the
components of a facility [9].Later that, Echeverry,
et al. (1991) proposed formalization of sequencing
knowledge for construction scheduling, and
developed a knowledge based prototype system
which incorporates the same principles [7].
Cherneff, et al. (1991) introduced the BUILDER
Figure1. The KBES generic architecture. which integrates the CAD with schedule
generation. It is based on object oriented
III. THE ROLE OF AI IN PROJECT programming [11, 12], and consists of two main
SCHEDULING knowledge modules (KMs); first “Draw” for
creating and interpreting architectural floor plans
Developing a plan can be automated by deploying and second “Planner” for producing construction
schedule.

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 406


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

projeect design in BIM format, Geographica


G al
Bjork (19944) presentted researcch details of
B
Inforrmation Syystem (GIIS), Weath her recordds
RAATAS project which concerns developingg an
databbase and Coontext. The system imp plementatioon
inffrastructure for computter integrateed construcction
invollves buildding consistent com mponents to t
annd digital transfer of inform mation am mong
mainntain interacction between these componentts
applications [13].
and provide
p fast and quick aaccess to eaach other.
R
Researchers at the Centrre of Integrrated Facilitty at
Revit provides API interroperability that allow ws
Staanford Unniversity have h carrieed out many
m
accesss and filterr elements of a design n model. On O
sciientific ressearch worrks for auutomating and
the other hand, the syystem gets access to t
seqquencing of constrruction activities. They T
Primaavera databbase, Primavvera P6 stores data in a
developing a knowleddge based environments
moree than 170 Tables.
T These tables arre referenceed
plaatform for project
p scheeduling.
to eaach other byb relationss between primary
p keey
Fischer andd Alami (1996) ( monstrated an and foreign
dem f keyy. There aree only arou und 20 mosst
emmpirical schheduling appproach thaat is basedd on used tables that anyone neeeds really to o be familiaar
computer models
m thatt simulatee interpretaable with.
construction method
m [11,, 12].
KBE ES Components for p planning prrojects: Thhe
KBES needs thhree modulees to perforrm planninng
IVV. AI AND D EXPERT SCHEDUL LING and schedulingg processees. They are namelly
Context, Operaators and K Knowledgee base. Thhe
Knnowledge representati
r on, reasoniing techniqques Cont ext Modul le contains s information on thhe
derived from m AI permiit computers to geneerate projeect, includinng the desiign, site ch haracteristiccs
plaans, not merely
m anaalyse plans produced by and the
t planninng decision
ns made. T he Operatoor
huumans. The knowleddge to geenerate pllans, Modu ule containns operators s that creatte, delete oro
descriptions of actionss, and conntrol struccture modiify the infformation sstored in the t contexxt.
reppresented foor selectingg new actionns to insert into Operrators are u
used for d
different tas
sks such a
as
a project
p plann. techn
nology chooice, activiity synthessis, duratioon
estim
mation ...etcc. The geneeric compo onents of AI A
In this paper an experieence modellling for hoouse application in KBES K and thheir relation
nship areas in i
buuilding projeect planningg is presenteed. It is useed to Figurre 2.
plaan this project, desccribe its performance
p e in
plaanning the constructionn of large‐sscale buildinngs.
D
Domain Experrt E
End User
Thhe Revit BIM B packkage and Primavera
P are
intterfaced to plan this project. Revvit can geneerate
loggically corrrect activitty networkks for buildding
construction from
f t components
a desccription of the Expert Syystem
off a facility, while Prim mavera maanage the tiime-
cost schedule for activitiies. To moddel projects in a
unniform fram mework, we will show the usefulnness
K
Knoledge Enginneer Building Tools
off some princciples for esstablishing relationship
r ps of
thee project componeents invollved in the Figu ure2. Main components
c s of KBES.
construction activities.
a
T AI/KBE
The ES application requirees developm ment The Knowledge-Base containss distincct
off a knowleddge base that t permits engineerss to knowwledge souurces of ta
ables, tree
es, semanti ic
intteract witth: the characteristi
c ic data, the netwo orks andd rulesspeecific to particulaar
theeoretical orr simulationn results, annd the speccific techn
nology choiices, activitty durations, and otheer
hyypothesis off control measures. KB BES have very v consiiderations.
higgh potentiaal for tackkling problems that lack l
exxplicit algorrithms (e.gg., problems for whicch a V. PR ROGRAM MMING, IM MPLEMENTATION
nuumerical moodel does not n exists).Itt retrieves data
d
froom many tyypes of docuuments, i.e. the producct or The C#.net proogramming language program p waas

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 407


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

used for development in this work. The object filled and there is no missing information in these
oriented properties of C# are explored by records.
modelling each element in the project design into a The KBES algorithm builds a database entity for
class that implicates the data structure needed for each element and defines the basic activities in
construction project element. The operators in entity relationships by expanding abstracted
KBES and the design activities used in planning design elements relationship to multiple levels of
and scheduling phases are structured and details as in figure 3. As noted earlier, the system
programmed as functions. defines frames for each design element. Each
The program starts by instantiating an operator Frame contains information concerning planning
which reads all elements from design model and process, then the system translates the Design
converts them to frames (design frames) stored in element to Element activity and then to Project
context space. After creating the frames, another activity as shown in figure 4.The KBES platform
group of operators that handles these frames by consists of primary tasks; inputs, outputs,
selecting construction methods, and aggregating or constrains/controls functions, and mechanisms.
splitting them according to the knowledge The inference engine uses the knowledge to
inferring. Before aggregation or splitting, some construct the line of reasoning that finds a solution
operators invoked to exam the design frame if it is for the construction problem. The KBES inference
not primitive (i.e. it is required another action to engine works in conversational mode, i.e. interact
be completed), the operator injects new frame/s with the user by questions, then the engine will
(i.e. Activity frame) which represent the extra work on a set of the reasoning rules to solve the
work needed to accomplish design element. The problem and achieve the goal gradually.
fields of each frame continue to be filled by
multiple operators till the frame records are fully

Simple House

Wall 2
Host
Wall 1 Door
Window

Inner Layer Part of Outer Layer

Core Layer
Inferred Relation

Part of

Reinforcing Framing Pouring Curing Removing Frame

Figure 3. The Relationship Expansion Design diagram.

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 408


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

Figure 4. Translattion of the Design


D Elem
ments into Construction
C n Activity

.
VI. EX
XPERIMEN
NTATION
N RESULTS
S steeps in creatting a consstruction pro
oject scheddule
caan be eliminnated. It alsso proved th
hat translation
A simple house buildding, as shoown in figurre 5, is off data bettween the Revit an nd Primavvera
used as a case study for validatting the prooposed so
oftware is automated
a to minimizze the hum man
system perrformance. It is an exaample for teesting, errrors.
investigatiing the functions correctnesss of
Element Extractor Operator, evaluatingg the
VII. CONC
CLUSION
results of the knowlledge base,, and finallly for
conductingg and dem monstrating interactionn with
In this paperr, a framew work for in ntelligent BIM
Primaveraa. This step is done byy correctly filling
dataa representaation and m
methods for use in projject
interfacingg tables. Eventually,
E the 3-dimeension
plannning were introduced. The archittecture and the
house is designed
d onn Revit sofftware, thenn each
function of KB BES for coonstruction planning was w
element coonstruction parameterss are passedd into
described. Appplying the K KBES repreesentations for
the Primavera softtware as input dataa for
each h stage off project pplanning an nd scheduling
constructioon activity schedulingg. The produced
dem
monstrated thet feasibillity of the system
s in that
t
data of the abovve activitiies are loaded l
actiivities netw
work, in bothh the durations estimatted,
automaticaally throughh the interffacing tablees into
and
d project schhedule.
the primavvera fields shown in figure 6 foor the
next step in scheduliing and maanagement of o the A number
n of feeatures are nnotable for the system;;
project acttivities. Thee proposed steps proveed that Firsst, the devveloped syystem led to a form mal
some of thet most teedious and time consuuming description off the construction pro oject planning

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 409


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

pprocess. It may be onne of the first knowlledge-


b
based systems that cann handle/ help
h construuction Second, the arrchitecture of the systeem is modullar,
p
planning m model dessign to scheeduled sem
prrocess from mi automaticc which maakes it flexiible and eassily
p
plan. app
plicable for other
o types of industriaal projects.

Figuure 5 Sam
mple Parameetric Window
w for buildding (case sstudy).

The propoosed system incorpoorated num merous VIII. REFERE


ENCES
k
knowledge sources to perform tecchnology choice, [1] Alami, F. B., Fischer, M. A., and Kun nz, J. C. (19998).
d
duration e
estimation, precedencce setting and "AEEC 4D produuction model:: Definition n and automaated
a
activity idenntification. geneeration", CIFE
E Working Papper #52. Stanfford Universitty.
The hieraarchy of prroject activvities proviides a
d
detailed work breakddown but also represents [2] Associated General Conntractors of America
A (AG
GC),
a
appropriate aggregatioons of activvities alongg with 20066. "The Conntractor’s Guuide to Build
ding Informattion
r
resource avvailability and technoology choiice at Mod deling", AGC,, Arlington, V
Virginia.
a
appropriate levels of agggregation. [3] Alami, F. B., Kunz, J. C.., and Fischerr, M. A. (19998).
While the feasibility of an autoomated plaanning "Moodel-based mechanisms for automated activvity
sequ
uencing", Worrking Paper N
No. 50. Stanforrd University.
s
system has been demoonstrated, thhe desirabillity of
a expert system of this sort is still an open
an [4] Bjork, B. (1994).
( "RATAS projectt developing an
q
question. C
Considerably y more expperience witth the infraastructure for computer-inttegrated consttruction". Jourrnal
of Computing
C in Civil
C Eng., 8(44), 401-419.
s
system willl be required, especiaally field teesting.
H
However, t
the potentiial benefitss of the system [5]C
Cherneff, J., Logcher, R R., and Sriraam, D. (19991).
"Inteegrating CAD w
with consttruction-scheddule
c
could be subbstantial. geneeration".Journnal of Computting in Civil Engineering,
E 5
5(1),
pp 64-84.
6

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 410


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

[6] Eastman, C., Teicholz, P., Sacks, R., and Liston, L. Civil Engineering, Vol. 4, No. 2, 1990, pp. 155-176.
(2008). "BIM handbook", second edition, Hoboken, New [13]Zozaya-GorostizeC.,Hendrickson C., Rehak D. . (1988),
Jersey: John Wiley & Sons, Inc. "Proceedings of the 5th ISARC", Tokyo, Japan, PP; 511-520.
[7]Echeverry, D., Ibbs, C. W., Kim, S. (1991). "Sequencing [14]ChengM.-Y., CuanL. Li-,. Tsai H.-C, and C. Pi-Hung,
knowledge for construction scheduling". Journal of (2012) “Artificial Intelligence Approaches to Dynamic
Construction Engineeringand Management, 117(1), 118-30. Project Success Assessment Taxonomic,” Life Sci. J.
[8] Fischer, M., and Aalami, F. (1996). "Scheduling with [15] Piraquive F. N. D., Crespo R. G., and. García V. H. M,
computer-interpretable construction method models". Journal (2015) “Analysis and Improvement of the Management of IT
of Construction Engineering & Management, 122(4), 337-47. Projects,” IEEE Lat. Am. Trans., vol. 13, no. 7, pp. 2366–
[9] Hutchings, J. F. (2004). "Project scheduling handbook", 2371, 2015.
New York: Marcel Dekker, Inc.
[16] Daniel M. M., Juan C. F., (2015)” Artificial Intelligence
[10] Karta, N. A., and Levitt, R. E., (1990). "Intelligent applied to project success: a literature review” International
planning of construction projects" Journal of Computing in Journal of Artificial Intelligence and Interactive Multimedia,
Civil Engineering, 4(2), 155-176. Vol. 3, Nº5.
[11]Navinchandra, D., Sriram, D., and Logcher, R. D. [17]Dossich C. S., and Neff G., (2011)“Messy talk and clean
(1988). "Ghost: Project network generator". Journal of technology: communication, problem-solving and
Computing in Civil Engineering, 2(3), 239-254. collaboration using Building Information Modelling”,
Engineering Project Organization Journal, 1(2), pp 83-93.
[12] Nabil A. Kartam, Raymond E. Levitt, "Intelligent
Planning of Construction Projects" Journal of Computing in

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 411


2019 16th International Multi-Conference on Systems, Signals & Devices (SSD'19)

Figure 6. The Primavera Results Window of the Case Study in figure 4.

978-1-7281-1820-8/19/$31.00 ©2019 IEEE 412

View publication stats

You might also like