Intelligent Agent-Based Subcontracting System in Construction

Practically, the effective management of subcontracting selection within a construction project has been regarded as one of the critical factors for achieving project success. This study takes advantage of IT initiatives of e-commerce to combine with the use of electronic information exchange standards Java XML (extensible markup language) to propose a “push-strategic” and agent-based online subcontract bidding and negotiation architecture. The design of agent will make it possible to automatically sense the changes in its environment and react accordingly on behalf of the host. The information agent deals with all manipulation of incoming and outgoing messages following the pre-defined communication mechanism. Furthermore, it will communicate with other information agents mounted on all other contract nodes, and also with its local decision support system through the mapping rules. The ultimate goal of this study is to build an agent-based automatic subcontracting environment for Multi-Agent System (MAS) to achieve the effective communication and optimization within the process of subcontractor selection. A proto information system has been designed and implemented in the study. The proposed agent-based subcontracting system, after being made into a widespread and workable process in the future, could expectably become a new paradigm for the subcontracting procurement of the construction industry.