EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

1713. Job Scheduling by Choosing and Tiling Existing Scheduling Segments

Invited abstract in session WC-49: Machine scheduling problems, stream Lot Sizing, Lot Scheduling and Production Planning.

Wednesday, 12:30-14:00
Room: M1 (building: 101)

Authors (first author is the speaker)

1. Gavriel David Pinto
Industrial Engineering and Management, Azrieli College of Engineering Jerusalem
2. Inessa Ainbinder
Industrial Engineering & Management, Azrieli Academic College of Engineering
3. Yehuda Hassin
Software Engineering, Azrieli College of Engineering Jerusalem
4. Daniel Lifshitz
Ben Gurion University of the Negev
5. Uriel Israeli
Faculty of Industrial Engineering and Technology Management, Hit- holon institute of technology
6. Gad Rabinowitz
Industrial Engineering & Management, Ben-Gurion University

Abstract

In a flexible job shop problem, a typical job consists of a set of operations with precedence relationships among them. Each operation may be performed in various modes, where each mode requires certain renewable resources, each for some specific time-period within the duration of the operation. We call the resources’ timings for performing a specific operation in a specific mode a scheduling segments. We assume that scheduling experts develop and keep for reuse a library of alternative scheduling segments for each operation. The reuse of a library of scheduling segments is of much importance for the "engineer to order" process where a manufacturer meets the specifications of their customer by engineering and producing the product after an order has been received. Developing a scheduling segment is a complex and time-consuming task, for which scheduling experts can benefit by reusing existing scheduling segments. The objective is to minimize the makespan by combining scheduling segments into a feasible schedule. We propose a new efficient heuristic for addressing this challenge. Previous methods for optimally solving this problem suffered from severe scalability issues, while previous heuristics experienced high rates of infeasibility that hurt both efficiency and performance. The proposed approach integrates a genetic algorithm with a greedy procedure for overcoming these difficulties while enabling, for the first time, the solution of practically sized jobs.

Keywords

Status: accepted

Back to the list of papers