Deep learning approach for automated guided vehicle system

Abstract

Automated guided vehicles system (AGVS) is a new logistics problem area, and the most demanded in terms of performance in view of the exponential growth of product traffic in the world. The present paper aims to to propose a new approach, to deal with AGVS problems, based on deep reinforcement learning algorithms, as an alternative to classic methods. Indeed, the classical approaches are based on a pre-established policy of vehicle movement rules, while our method deduces the movement rules based on trial/error reinforcement learning approach, and effectively gives good results in relatively small moving areas and a limited number of agents.