A multi-agent design for sense and respond logistics simulation

Abstract

Sense and Respond Logistics (SARL) is a highly flexible, agile, and adaptive methodology for managing logistics resources in a network-centric and information rich environment. Designing and developing the software for simulating a system with such complexity presents challenges that stem from the role of heterogeneous communication channels and decision choices that produce non-deterministic outcomes. We describe the use of multi-agent approaches for simulating SARL applications. We also argue for the use of formal methods and aspect-oriented techniques in the design of these simulations, to assist in supporting model validation and verification. Features of an aspect-oriented, multi-agent, architectural description languages are also presented as a mechanism for reasoning about cross-cutting concerns.