Scheduling for deadlock avoidance operation in robotic manufacturing cells

Abstract

This paper addresses a scheduling and control method for deadlock avoidance operation in robotic manufacturing cells with flexible job sequences. The proposed scheduling method generates an optimal or near-optimal schedule avoiding deadlock even when an unexpected event such as machine failure happens. The proposed deadlock-free scheduling approach consists of two components: deadlock manager, and dynamic scheduler. The deadlock manager builds and manages deadlock in a robotic manufacturing cell. The resources or machines in the robotic manufacturing cell are classified into two categories: potential deadlock set and deadlock-safe set. This paper presents an efficient method to obtain a minimal potential deadlock set using a genetic algorithm. Scheduling with deadlock avoidance policy is applied to the potential deadlock set, and scheduling only is applied to the deadlock-safe set. The dynamic scheduler generates a deadlock-free schedule using piecewise shifting operations in a real-time operation phase. The proposed scheduling approach is shown by simulation to be efficient in managing deadlock inherent to the robotic manufacturing cells.