A Novel Framework for Automated Warehouse Layout Generation

TL;DR

The paper tackles automated warehouse layout generation by framing it as a constrained optimization problem that balances storage capacity and accessibility. It introduces a novel candidate-layout generator based on constrained beam search over a discretized space, coupled with Layout Filtering, a bespoke Layout Scoring function, and a Connectivity metric to rank high-throughput candidates. A post-refinement step tailors layouts to site-specific constraints, and implementation leverages multiprocessing to scale the search. Empirical results on real-space cases show Pareto-optimal layouts that outperform manual designs and are validated with expert designers, offering a practical tool for rapid exploration and decision-making in warehouse design.

Abstract

Optimizing warehouse layouts is crucial due to its significant impact on efficiency and productivity. We present an AI-driven framework for automated warehouse layout generation. This framework employs constrained beam search to derive optimal layouts within given spatial parameters, adhering to all functional requirements. The feasibility of the generated layouts is verified based on criteria such as item accessibility, required minimum clearances, and aisle connectivity. A scoring function is then used to evaluate the feasible layouts considering the number of storage locations, access points, and accessibility costs. We demonstrate our method's ability to produce feasible, optimal layouts for a variety of warehouse dimensions and shapes, diverse door placements, and interconnections. This approach, currently being prepared for deployment, will enable human designers to rapidly explore and confirm options, facilitating the selection of the most appropriate layout for their use-case.

Figures (2)

• Figure 1: Running example of beam search ($b=1$) for a sample space from our industry partner with specified door connections and aisle width = 3. At the initial state, all free spaces are assigned as storage spaces. At the first level, all possible unique children are generated by carving aisles with sliding horizontal and vertical aisles across all block stores. Depending on the particular $\alpha-\theta$ combination specified, a different configuration is found as the best solution for that setting. This carving process is undertaken again on the chosen configuration in subsequent levels until a terminal state is found. In the figure, each colored path represents a route that leads to the best possible solution for a particular $\alpha-\theta$ combination. For each of these solutions, the $\alpha$, $\theta$, number of pick faces and number of storage are shown. Space specifications: Walls, Door connections, Aisles, Storage, Pick face.
• Figure 2: Pareto visualization for a medium-sized space from our industry partner with specified door connections and aisle width = 3. The pink dashed line shows the Pareto front. Zoomed layouts correspond to the data points on the Pareto front. The red star at the bottom shows the manually-designed layout which has been specified by the red dashed border. Space specifications include the following: Walls, Door connections, Aisles, Storage, Pick face.