2361. Warehouse-Aware Design and Evaluation of Algorithmic Pipelines for Decision-Making in Warehouse Operations
Invited abstract in session WB-12: AI and Optimization for Warehousing, stream Artificial Intelligence, Machine Learning and Optimization.
Wednesday, 10:45-12:15Room: H10
Authors (first author is the speaker)
| 1. | Janik Bischoff
|
| 2. | Hadi Kutabi
|
| Karlsruhe Institute of Technology | |
| 3. | Maximilian Barlang
|
| Institute for Material Handling and Logistics (IFL), Karlsruhe Institute of Technology | |
| 4. | Özge Nur Subas
|
| Operations Research, Karlsruhe Institute of Technology | |
| 5. | Uta Mohring
|
| Eindhoven University of Technology | |
| 6. | Fabian Dunke
|
| Institute for Operations Research (IOR), Karlsruhe Institute of Technology (KIT) | |
| 7. | Anne Meyer
|
| Karlsruhe Institute of Technology | |
| 8. | Stefan Nickel
|
| Institute for Operations Research (IOR), Karlsruhe Institute of Technology (KIT) | |
| 9. | Kai Furmans
|
| Karlsruhe Institute of Technology |
Abstract
Optimization problems in warehouse operations---such as picker routing, order batching, and picker sequencing---are typically addressed sequentially rather than in an integrated manner, due to their high computational complexity.
However, the quality of the overall solution produced by sequentially executed decision-making processes in warehouse operations depends on the selected combination of algorithms for the respective subproblems.
Furthermore, it is not apparent which algorithms are applicable given a certain warehouse system or instance due to the missing semantic modeling of the warehouse context and algorithm requirements.
We aim to close this gap by developing a framework that can provide and evaluate all feasible algorithm combinations for a given warehouse system.
The proposed methodology includes:
(1) the semantic description of the warehouse system and required data;
(2) mapping between the warehouse system and compatible algorithmic approaches;
(3) design, automated synthesis, and efficient execution of algorithmic pipelines;
and (4) selection of the best pipeline based on a comprehensive performance evaluation.
To showcase the effectiveness of our approach, we apply it to nine benchmark instance sets from the literature, covering the problem classes of order picking, joint order picking and batching, and picker sequencing.
Consequently, this work contributes to the warehouse optimization literature by providing a generalizable framework for the design, automated synthesis, and selection of valid and well-performing algorithmic pipelines based on a warehouse context.
Keywords
- Decision Support Systems
- Warehouse Design, Planning, and Control
- Logistics
Status: accepted
Back to the list of papers