Scheduling considers the problem of allocating scarce resources to activities over time. It plays an important role in delivering projects on time and within budget. However, it has long been a challenge to housing construction domain because of the intrinsically distributed project delivery practice. Moreover, with the emergent customer requirements on customization, housing construction project scheduling are confronted with larger variation in tasks, higher dynamics in tasks arrival, tighter project delivery time window, under which, traditional centralized scheduling approaches are no longer efficient. Therefore, it is desirable to have a scheduling mechanism that takes into account the distributed nature of construction scheduling problem and incorporates temporal dynamics arisen f...[ Read more ]

Scheduling considers the problem of allocating scarce resources to activities over time. It plays an important role in delivering projects on time and within budget. However, it has long been a challenge to housing construction domain because of the intrinsically distributed project delivery practice. Moreover, with the emergent customer requirements on customization, housing construction project scheduling are confronted with larger variation in tasks, higher dynamics in tasks arrival, tighter project delivery time window, under which, traditional centralized scheduling approaches are no longer efficient. Therefore, it is desirable to have a scheduling mechanism that takes into account the distributed nature of construction scheduling problem and incorporates temporal dynamics arisen from customization.

Multi-agent system intrinsically provides the flexibility, robustness and reactivity in distributed problem solving instances. We explore, in this thesis, the feasibility and issued on applying agent-based scheduling mechanism into construction scheduling context. We tackle the decentralized scheduling problem from resource point of view and propose an auction based negotiation protocol for coordination among distributed scheduling agents. We evaluate the performance of propose scheduling mechanism under different activities load and different sets of bidding strategies. The experimental results show that our proposed fully distributed scheduling mechanism is able to reach satisfactory resource leveling under different scenarios.