Automating nurse self-rostering: A multiagent systems model

This paper presents a multiagent systems model for constructing nurse rosters. Over the past decade self-rostering has become more favorable in nursing personnel scheduling, due to its empowerment and motivational benefits. However, the labor intensive negotiation procedure among participants has limited its application to medium sized and large wards. To overcome this limitation, we propose an automated negotiation tool utilizing economic-based negotiation mechanisms. We model the environment as a multiagent system. Nurses can indicate their preferences by configuring the preference profiles of the nurse agents, and the nurse agents collectively construct rosters through the negotiation with the management agents and among themselves. To support the design and implementation of automated self-rostering systems, we present a multiagent systems architecture, a structure of preference information, and a negotiation mechanism.