A Decision Support System for Pre-Cast Concrete Manufacturing Planning: An Innovative Crew Allocation Optimiser

A Decision Support System for Pre-Cast Concrete Manufacturing Planning: An Innovative Crew Allocation Optimiser
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  CSCE 2009 Annual General Conference Congrès annuel générale annuelle SCGC 2009  St. John’s, Newfoundland and Labr  ador /  St. John’s, Terre -Neuve et Labrador   May 27-30, 2009 /  27-30 mai 2009  1 A Decision Support System for Pre-Cast Concrete ManufacturingPlanning: An Innovative Crew Allocation Optimiser Ammar Al-Bazi 1 , and Nashwan Dawood 21 PhD Researcher, Center for Construction Innovation Research (CCIR), School of Science andTechnology (SST), University of Teesside, Middlesbrough, TS1 3BA, UK, Tel. +44 1642 342 494, Email:    2 Professor ,Center for Construction Innovation Research (CCIR), School of Science and Technology(SST), University of Teesside, Middlesbrough, TS1 3BA, UK, Tel. +44 1642 342 494, Email:    Abstract: The high cost of skilled labour in the precast concrete industry and the dynamic nature of theproduction processes have encouraged senior managers in the industry to develop a more intelligent andoptimal allocation strategies. In this paper, the Genetic Algorithm (GA)-based simulation optimisationapproach is used for the optimal allocation of different crews of workers on different precast concreteproduction processes. A Precast-Component manufacturing simulation model is integrated to GA-basedoptimisation model to simulate the physical processes that are involved in the manufacturing system andto optimise the allocation of labour crews. The proposed approach determines the optimal or near optimalallocation of crews for the processes involved within the precast manufacturing system., This shouldeventually lead to maximum utilisation of the set of skilled workers involved in the allocated crew andsubsequently minimise the total labour costs. The paper discusses a simulation system dubbed “SIM_Crew” that was developed during this study. Thissimulation model is developed as a test bench for the proposed intelligent allocation system. GA is usedto guide simulation towards the best course of action; chromosome is designed well to involve all thedecision variables. An innovative selection procedure is developed in order to guarantee the bestselection of chromosomes. The results showed that efficient utilization of skilled labour had a substantialimpact on reducing the throughput time, minimising labour costs, idle times and maximising the skilledworkers utilisation. 1. Introduction The precast concrete sector in the UK produces over 36 million tones of products every year worth morethan £2.3 billion and employs in excess of 20,000 people. The widely predicted growth for the UKconstruction industry over the next few years presents an excellent opportunity for the precast sector, aswell as a significant challenge as the overall market of labour, both skilled and unskilled, become evenmore competitive beside the associated costs.   2Many of the manufacturing systems in such industry involve highly skilled workers and experiencedsupervisors to carryout the production processes. The labour intensive production processes are thoseprocesses that require high level of labour compared to capital investment. Such processes are morelikely to be seen in job shop production.The high cost of skilled labour in the precast industry is a force driving the precaster to improveproductivity and hence minimise the total production costs. To improve the productivity in a productionfacility, optimal allocation of resources and best planning of workforce are required. The best allocation ofresources will eventually lead to minimising wastages and guarantee best flow of work.The precasters are trying hard to achieve best allocation of resources on different processes. This is acomplex problem due to the vast amount of different possible solutions. Often this kind of problem iscalled a hard problem; and the classical solutions to these types of problems cannot be used to obtainsatisfactory results. More sophisticated tools are required to assist in the decision making processes ofresource allocation.The lack of using innovative tools to solve crew allocation problems in the precast manufacturing systemshave motivated this research to develop and rehearse an innovative allocation system that could assistprecasters as a decision support system in order to improve performance of their precast manufacturingsystems. The optimal allocation of crew will eventually minimise labour costs, maximise labourerutilisation, minimise process idle time and subsequently improve productivity of manufacturing systems inthe precast industry.The paper is structured as follows: in the next section, literature review concerns resource allocationtechniques is presented. In section 3, concept of the SIM_Crew is presented. Then materials andmethods, which have been used in designing the intelligent model, is discussed in details in section 4.Case study is demonstrated in section 5 followed by results in section 6. Conclusions are presented insection 7, future works are the contents of the last section. Resource Allocation Problem in the Precast Industry Logical dependencies and resources (e.g., crew of workers and equipment such as machines) requiredwhen allocating resources in the precast project. The formation of a crew may contain some sharedworkers which they are involving in more than process.The required crew for each process may be allocated individually or come may come from other activitiesthat share the same resource(s). The individual resource allocation allows activities to start as soon asthe logical dependencies (i.e., the completion of the preceding activities) are available, with no delay (idletime for the resources).Because both parallel and sequencing nature of some or all production processes of a project is pre-specified, the labour shared by multiple similar or different processes are generally allocated according tothe work flow path. However, idle times or resources are still not carefully considered and the overallthroughput time or total costs maybe longer or higher than expected. In fact, such a problem becomesmore significant when there is a significant idle time for a process(s) because of shared workers allocatedat more than one process and needed at each of allocated crews amongst different or similar precastoperations.Because of the precast operations are labour extensive and the workers involved in such industry arehighly skilled and costly, optimisation of the crews allocation amongst the precast manufacturing systemis needed in order to follow up each allocated worker on one or more than process to minimised the idletime caused by the shared usage of a worker. The minimisation of idle times subsequently will improveproductivity and efficiency of such manufacturing systems.   3 Figure 1: The problem of crew’s allocation on similar/different processesEach parallel processes have a certain pool of crews that can possibly perform the work for each of theparalled process, Each crew has different set of workers with a certain process time. The formation ofeach crew (set of workers involved) might contain shared labour. The shared labour available into morethan formation will eventually cause a delay in process(s).Next section will present the architecture of the SIM_Crew which is an intelligent allocation optimiser that has been proposed and developed to achieve best crew’s allocation to production processes. This allocation eventually will reduce clashes between any shared resources, and subsequently lead tominimising resources cost and maximising utilisation of used resources. 2. Literature Review of Crew Allocation Several research projects using innovative tools to improve the performance of the precast concreteproduct manufacturing systems have been conducted. Murphy, G. (2001), combined tabu searchprinciples with a simple improvement- swapping heuristic to allocate stands and cutting patterns tologging crews for a single time period. Two comparisons were made between the TABU heuristic (500iterations) and an IP formulation for a small problem (10 stands, 5 logging crews, 5 cutting patterns, and 5log types). The first test allowed as many logging crews to be located in any one stand as was required.The tabu search solution was within 0.8% of the IP solution. The second test allowed a maximum of onecrew per stand.Guttkuhn R. et al (2003) introduced a discrete event simulation for crew assignment and crew movementsbecause of train traffic, labour rules, government regulations and optional crew schedules. The systemhelps to evaluate changes to current crew assignments and can test new crew assignment scenariossuch as crew schedules. The system is also used to assess the impact of traffic changes on existing crewschedules in order to implement reactive corrections to these schedules. The results of the simulationallow the user to draw conclusions concerning the operational characteristic of the real crew assignmentprocess.   4Moselhi, O. et. al (2006) proposed a new methodology utilises and combined Genetic Algorithums andspatial technologies for optimisation of crew formations for earthmoving operations. The designed modelis useful for planning, tracking and control of earthmoving operations. Genetic Algorithum is used inconjunction with a set of rules developed to speedup the optimisation process and avoid generating andevaluating hypothetical and unrealistic crew formations. The model accounts for resources that areavailable to contractors and it is capable of reconfiguring crew formation dynamically during theconstruction phase while site operations are in progress. The results have shown that the use of spatialtechnology (GPS) for data collection provides project teams with timely, inexpensive, and accuratemonitoring tool. In addition, without using of waiting time rule, GA would have selected unrealistic crewformation although it has the minimum possible cost.Li, H. et al (1998) presented a methodology for optimising labour and equipment assignment forexcavation and earth work tasks using the Genetic Algorithum. A number of modifications to features ofthe basic genetic algorithms are conducted to improve the capacity of the genetic algorithms. Five GAruns for the labour and equipment assignment problem were performed, each with twenty differentsolutions. The results were shown that the lowest cost solution is achieved in run 1. The other runs allachieved near-optimal solutions with difference from the optimum in the range of 0.47%.Agbulos, A. et al (2003) developed a simulation analysis based lean thinking to improve the performanceof drainage operations maintenance activities. Application of lean concepts to drainage operationsmaintenance crews relied on assumptions due to either poor quality data or data not collected al all. Theintegration of lean theory and simulation methodology provides the tools necessary to improve labourproductivity for drainage operations maintenance activities.Matt (2007) used agent based modelling methods to simulate space congestion on a construction site toexplore the impacts of individual interactions on productivity and labour flow. In the agent-based model,each worker and task is represented as an autonomous agent. A simulation for two masonry crewsintersecting in space is conducted. Each crew consists of 3 brick layer agents and 2 helpers.Thesimulation was intended to study the effects of the sizes of labour crews on productivity. In the currentimplementation, the skilled workers proceed toward the wall, and complete their nearest tasks.Zhang H., et al (2004) developed an optimisation methodology which integrates discrete-event simulationwith a heuristic algorithum to optimise dynamic resource allocation for construction scheduling. Threekinds of heuristic al location policies for the crew were considered, “ fixed  ”, “ first-use  ” and the “ heuristic  ” allocation policies. The utilisation rates of each group of crew for different allocation polices arecompared. As a result, the optimal allocation leads to the highest utilisation rates of the four groups of crew, the “ first-first  ” policy generates the average once, and “ fixed  ” allocation policy produces the lowest ones.In a summary, the reviewed literatures have revealed that there is a lack of knowledge of usingsophisticated techniques as simulation optimisation to improve performance and reliability of precastconcrete products processes by achieving optimal allocation of labour in such industry. The primaryobjective of the study was to use simulation modelling to analyse effects of resource allocation onimproving performance of the precast concrete manufacturing system. Computer simulation model wasdeveloped using ARENA Rockwell Simulation software to simulate the tasks or processes that operatorsperform when producing precast concrete products.   5 3. Concept of the SIM_Crew The purpose of allocating crews is to place the right and most suitable team of workers for each processat minimum associated cost and for better performance. Therefore, as a fundamental requirement, crewsshould be allocated on processes without any tardiness. In addition, performance of alternative feasiblecrews should be evaluated in terms of total scheduling-related cost so that the one with the minimum costand interruption is selected for implementation.To meet the above allocation process objective, a Simulation-GA based model is proposed, as shown infigure 1.Figure 1: SIM_Crew Intelligent System ArchitectureIn this model, possible alternatives of crews were allocated on different processes in order to come upwith the best allocation at minimum cost, reduced idle time for production processes and maximumutilisation of workers. Thus the decision variables include the set of possible and feasible crews for everyprocess with different set of workers involved at each crew.As shown in figure 1, the allocation process is an iterative procedure of progressive improvement. Duringone allocation iteration, simulation excutes allocation plans: each of them has sets of crews to beallocated on production processes, while GA evaluates the performance of the resultant allocation, andbased on this, adjusts the decision variables and selects the most promising solution. 4. Process Mapping, Simulation and Genetic Algorithms To conduct a structural analysis for the processes involved in the precast manufacturing system, IDEF0diagrams were used to apply structural methods to better understand how to improve manufacturingproductivity.The data collection and process modeling is divided into three main phases: Data collection, processmodeling, documentation, and the action phases.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks