Final Report Summary - BUILD-CYCLE (Integrated analysis and visualization for advanced construction planning and controls)
Recent technological developments, such as building information models (BIM) promise major construction productivity gains, supporting design-construction integration and improvements in virtually all functions. However, there is still a lack of effective process analysis tools for the construction stage using BIM. Model-based, detailed, and realistic analysis of construction projects is necessary to better evaluate alternatives and to make high impact decisions whilst keeping the overall perspective of the project.
BUILD-CYCLE project aims to improve the construction process analysis by providing a general purpose resource-integrated simulation platform on 3D building information models. Moreover, it aims to make this analysis easy and extensible with parametric definition of the process. To that end, it combines several new approaches.
One of the major roles of construction management on site is defining and orchestrating resources. Labour crews use various tools, equipment, and materials to perform the work. Site supervision directs crews and equipment to perform tasks in a specific sequence and at a specific target production rate. To reflect this type of construction resources, we built an agent-based simulation model for resources, where each individual crew and major equipment is an agent with specific goals and policies. By providing simple and natural policies for resources, realistic construction practices can be achieved.
With the aid of geometric techniques, we abstract spatially complex work sequences of multiple activities to complete building element locations on BIM. We analyse this construction model on BIM through a hybrid simulation approach, where discrete-event simulation model for activity sequences are unified with the agent-based simulation model for resources.
We have developed prototype software called GSimX from scratch as part of the BUILD-CYCLE project to implement this modelling and simulation approach. In this 3D platform, planners can import data from BIM and project schedule data from common software packages, and interactively define resources and their behaviour. Users can generate method, sequence, and resource alternatives, and evaluate performance, duration and cost results as a combination of summary values, graphs, and 3D visualization. The platform supports extensions to define new equipment and crews, update resource behaviour and support new construction scenarios.
Using this approach for integrated construction planning and controls can benefit various areas of construction and we have focused on two primary areas. The first is pre-construction alternative analysis. There is no well-established automation approach for early stage resource and sequence alternative evaluation for construction. Evaluating many alternatives for early project decisions such as the type and quantity of major equipment, crew allocation and work sequencing, has important implications for the rest of the project. Through realistic crew and equipment support, accurate cost representation, and comparison charts, we have demonstrated that multiple pre-construction alternatives can easily be evaluated with GSimX. With this approach, planners can make better resource and sequence decisions before developing schedules and estimating the project in detail.
The second application area is resource and BIM integrated short-term construction planning. The disconnect between different levels of construction plans remains a major issue limiting automation in construction. Complexity of information coming from various participants and the site to perform such analysis usually leads to manual planning approaches, which leads to wasted potential for optimizations and challenges for project controls and re-planning. The researcher has analyzed requirements for this platform to support look-ahead planning and short-term planning following lean construction principles. This study has resulted in a workflow that supports generating look-ahead and short term plans and re-planning based on progress data transferred from detailed levels to the master level. The approach enables detailed project control during construction, frequently updating the plan through progress monitoring, tracking resource use, and perform forecasts and updates as necessary.
This overall research project is a step towards a general framework for planning and control implementation for construction projects. Model-based construction process analysis will lead to a more efficient use of project resources. This will improve not only the timing and cost of the project, but also energy consumption and greenhouse gas emissions for the construction phase. Furthermore, this platform is expected to enable application of artificial intelligence techniques to improve automation for construction processes.
BUILD-CYCLE project aims to improve the construction process analysis by providing a general purpose resource-integrated simulation platform on 3D building information models. Moreover, it aims to make this analysis easy and extensible with parametric definition of the process. To that end, it combines several new approaches.
One of the major roles of construction management on site is defining and orchestrating resources. Labour crews use various tools, equipment, and materials to perform the work. Site supervision directs crews and equipment to perform tasks in a specific sequence and at a specific target production rate. To reflect this type of construction resources, we built an agent-based simulation model for resources, where each individual crew and major equipment is an agent with specific goals and policies. By providing simple and natural policies for resources, realistic construction practices can be achieved.
With the aid of geometric techniques, we abstract spatially complex work sequences of multiple activities to complete building element locations on BIM. We analyse this construction model on BIM through a hybrid simulation approach, where discrete-event simulation model for activity sequences are unified with the agent-based simulation model for resources.
We have developed prototype software called GSimX from scratch as part of the BUILD-CYCLE project to implement this modelling and simulation approach. In this 3D platform, planners can import data from BIM and project schedule data from common software packages, and interactively define resources and their behaviour. Users can generate method, sequence, and resource alternatives, and evaluate performance, duration and cost results as a combination of summary values, graphs, and 3D visualization. The platform supports extensions to define new equipment and crews, update resource behaviour and support new construction scenarios.
Using this approach for integrated construction planning and controls can benefit various areas of construction and we have focused on two primary areas. The first is pre-construction alternative analysis. There is no well-established automation approach for early stage resource and sequence alternative evaluation for construction. Evaluating many alternatives for early project decisions such as the type and quantity of major equipment, crew allocation and work sequencing, has important implications for the rest of the project. Through realistic crew and equipment support, accurate cost representation, and comparison charts, we have demonstrated that multiple pre-construction alternatives can easily be evaluated with GSimX. With this approach, planners can make better resource and sequence decisions before developing schedules and estimating the project in detail.
The second application area is resource and BIM integrated short-term construction planning. The disconnect between different levels of construction plans remains a major issue limiting automation in construction. Complexity of information coming from various participants and the site to perform such analysis usually leads to manual planning approaches, which leads to wasted potential for optimizations and challenges for project controls and re-planning. The researcher has analyzed requirements for this platform to support look-ahead planning and short-term planning following lean construction principles. This study has resulted in a workflow that supports generating look-ahead and short term plans and re-planning based on progress data transferred from detailed levels to the master level. The approach enables detailed project control during construction, frequently updating the plan through progress monitoring, tracking resource use, and perform forecasts and updates as necessary.
This overall research project is a step towards a general framework for planning and control implementation for construction projects. Model-based construction process analysis will lead to a more efficient use of project resources. This will improve not only the timing and cost of the project, but also energy consumption and greenhouse gas emissions for the construction phase. Furthermore, this platform is expected to enable application of artificial intelligence techniques to improve automation for construction processes.