A dynamic scheduling approach for optimizing the material handling operations in a robotic cell

This paper investigates a real-time dynamic job-shop scheduling problem in a robotic cell, in which multiple jobs enter into the cell with unexpected arriving rates. Different from classical flow-shop and job-shop scheduling problems, the jobs' transportation handled by a robot must be considered. Another characteristic is that the jobs' processing times are not constant values but confined in time-window constraints. To efficiently solve this problem in real time, the original schedule is restricted to zero changes. The problem is formulated as a sophisticated Mixed Integer Programming (MIP) model in which the new jobs' processing and transportation operations are inserted into the available time intervals of the original schedule. To strengthen the MIP model, speed-up constraints are added by taking advantage of specific relationships between the available time intervals arranged for a job's processing and transportation operations. Furthermore, an exact iterative algorithm is proposed, which starts with a relaxed solution of the MIP model and iteratively adds essential robot handling capacity constraints back to the relaxed MIP model until an optimal solution is found. Computational results validate effectiveness and efficiency of the strengthened MIP model and the iterative algorithm.