Abstract

As design complexity increases and clients demand faster deliveries, multiple activities involving machinery and labor are executed simultaneously, requiring the same space at the same time. This may lead to delays and safety risks. Previously developed models failed to acquire traction due to three gaps: lack of considering spatio-temporal uncertainties; absence of modelling the propagating effects of 4D-clashes on schedule progress; and not accounting for the unpredictability of crews’ decisions for resolving a 4D-clash. Additionally, the recent pandemic exacerbated vulnerabilities with the need to implement physical distancing, which created a new class of soft clashes. They are a result of different crews’ proximity to each other but do indicate a physical obstruction. This has triggered the need to introduce physical distance as a new independent workspace and to redefine the previously implemented clash detection rules.
Addressing the gaps, this thesis develops a framework and tool, adopting open BIM. Contributions are: simulating space as a probabilistic variable; automatically generating physical distance workspaces; smart clash detection identifying soft and operational clashes, while eliminating unrealistic and redundant results; automated clash resolution based on heuristic rules and constructability constraints; and modelling crews’ decision for handling unresolved clashes. The tool utilizes Blender’s GUI, benefiting from its open-source license and IfcOpenShell add-on. Additionally, an application using C# language was developed to specify user thresholds. 60 use-cases were created to test the functionality of the method/software. Comparing to manual calculations / geometrical intersection tests (used in clash detection software) the tool achieved a 99% accuracy in clash detection and volume estimation. The work was validated through two case studies. The first case study focused on labor workspaces for a residential project. The results showed a 50% clash (volume and instances) increase stochastic simulation without physical distances. The second case study simulated the refurbishment of an industrial plant during its shutdown. The solution presented 6,930 project completion possibilities recommending an 8% duration increase to achieve a 75% confidence level. The results were presented to industry experts whom acknowledged the benefits of the research in its current state, highlighting the ability to extend/adapt it to other domains.