A multi-agent system for decentralized multi-project scheduling with resource transfers

The companies dealing with multiple projects are geographically distributed at different locations. These projects require local (always available to the concerned project) and global (shared among the projects) resources that are available in limited quantity. The global resources are generally required to be transferred physically among the projects, consuming significant amounts of both time and cost. The existing multi-agent systems for decentralized resource constrained multi-project scheduling problem (DRCMPSP) do not consider the resource transfer time and related cost for execution and control. We introduce DRCMPSP-RT that explicitly considers the transfer of shared global resources. This paper proposes a novel distributed multi-agent system using auctions based negotiation (DMAS/RIA) approach for the resource intervals and allocating multiple different types of shared resources amongst multiple competing projects. To incorporate this feature the resources are planned to transfer in advance for the eligible successor activities to schedule within the current time. The proposed approach can solve complex large-sized multi-project instances without any limiting assumptions regarding the number of activities, shared resources or the number of projects. The DMAS/RIA is tested on standard set of 140 problem instances. Computational experiments compare the presented managerial approach with two transfer time neglecting approaches. The result found the relative GAP with an average of 33.43% with respect to average project delay (APD). Whereas, the percentage gap of total makespan (TMS) is obtained with an average of 16.52%. The result proves the impact of resource transfer time on project delay and increase in multi-project duration that can be minimized by project managers considering it in planning phase. (C) 2013 Elsevier B.V. All rights reserved.