A construction-specific simulation-based framework for earthworks.

Construction companies are operating within an increasingly competitive environment. Work often has to be tendered for on a very low profit basis. If the tender is too high, work is lost. If too low the contract may be won, but the job completed at a loss, unless more effective working methods can be found. Plans are used throughout the construction industry to allocate resources and schedule work. Yet, the planning tools used; Gantt chart, PERT and Queuing theory to name but a few, represent jobs as if they are static in duration, which in the complex, dynamic construction environment are clearly inappropriate. The EPSRC fuelled interest in developing a simulation methodology by suggesting that the construction industry could be considered similar to the traditional manufacturing industry. The manufacturing industry faced similar production dilemmas, work was completed but using inefficient resource configurations, causing bottlenecks, increased work-in-progress leading to higher costs. To reduce number of problems the manufacturing industry sought to utilise and develop a planning technique that had the capacity for modelling the dynamic nature of the industry. Discrete-event simulation enables the problems associated with manufacturing to be anticipated and minimised, as opposed to constantly fire-fighting. Since using simulation has accrued such impressive benefits within the manufacturing industry it is therefore not without credence to believe that the construction industry could also obtain saving from embracing this management tool.Simulation has been applied to model a number of scenarios within the construction industry. Similarities between the applications were sought and an area for further development was identified. A problem was modeled using the most frequently encountered simulation paradigms found in the manufacturing and construction industries, 'Activity cycle' and 'Process based'. Of the two methodologies, 'Process based' was selected for the development of further models. A conclusion drawn from the research is that simulation is not being utilized within the construction industry due to the perception that it requires an excessive use of resources. The research project identified that the model building process may be simplified through the development of generic simulation modules. These generic modules enable a simulation model to be developed quickly and easily by a non-simulation practitioner. The generic modules can be connected to represent the layout of an earthmoving operation. A host of scenarios can thus be modeled with the minimum of time and effort. To ensure that only significant data and process logic was collected and included within the modules the experimental methodology factor analysis was employed. Using this experimental technique, the relationship between and significance of ten different factors were established. Further experiments were performed on the most significant factors establishing an appropriate level of detail for those factors. It was beyond the scope of this thesis to develop modules for every conceivable construction process. Therefore, a methodology is given documenting the development of the chosen construction processes.