Incentive compatible, collaborative production scheduling with simple communication among distributed agents

Abstract

We study the collaborative production scheduling problem that arises when schedulers must coordinate their schedules with internal or external customers. We envision a setting where geographically dispersed schedulers (agents) coordinate their scheduling using asynchronous, web-enabled mechanisms. We propose the design of such a mechanism where resource allocation is communicated in the form of schedules over which each agent involved has different preferences and financial incentives. Specifically, we design a ‘schedule selection game’ where all participating agents state their preferences via a valuation scheme, and the mechanism selects a final schedule based on the collective input. We address the issue of incentive compatibility, recognizing that independent agents may not reveal their private information truthfully, and they may not behave in a way that is aligned with the overall system efficiency. Based on the principles of Vickrey–Groves–Clarke mechanisms, we show that the proposed scheme is a direct revelation mechanism that implements the optimal schedule selection under agents' dominant strategies. We analyse the main properties of the mechanism such as budget balancedness and individual rationality. We illustrate the operation of the mechanism using a real-world example in coordinating electronics component manufacturing.

Keywords:

• Production scheduling
• Incentive compatible
• Collaborative
• Distributed agents