Warehouse Order Batching for Route Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current order item picking strategies in warehouses are inefficient due to the need for significant computational power to determine optimal picking routes, especially when considering individual orders, and are less optimal over time, while also being topology-dependent.
Innovation Solution
A two-phase picking process is implemented, where orders are first batched based on calculated distances using a network representation of the storage facility, and then optimized using metaheuristics like simulated annealing or greedy randomized adaptive search procedures to determine efficient picking routes for each batch, reducing computational resources and accommodating various warehouse topologies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If optimal picking routes are determined based on individual orders using sequential filling and optimization, then route optimization is achieved, but computational power requirements increase significantly and solutions become less optimal over time
Solution Approach 1:
The patent segments the order set into multiple batches based on warehouse zone proximity. Instead of optimizing routes for all orders individually in sequence, the system divides orders into batches (e.g., Batch 1, Batch 2, Batch 3) where each batch contains orders with items located in specific warehouse zones. This segmentation reduces the computational complexity of route optimization for each batch while maintaining overall route quality.
Solution Approach 2:
The patent performs preliminary batching of orders based on distance metrics before route optimization. By pre-grouping orders into batches using distance-based clustering (calculating distances between order items and grouping nearby items), the system prepares the data structure in advance to enable more efficient subsequent route optimization with reduced computational requirements.
2Manufacturing precision
If comprehensive route optimization is performed for all orders, then optimal picking routes are achieved, but processing time increases
Solution Approach 1:
The patent divides the order processing into multiple batches, where each batch is optimized independently rather than optimizing all orders in a single comprehensive process. This segmentation of the optimization process reduces the time required for each optimization cycle while maintaining acceptable route quality for each batch.
Solution Approach 2:
The patent implements periodic batch processing where orders are processed in discrete batches rather than continuously optimizing all orders simultaneously. Each batch undergoes route optimization at periodic intervals, allowing the system to manage processing time effectively while still providing updated routing information as new orders arrive or existing orders are fulfilled.
3Productivity
If picking strategies are optimized for specific warehouse topologies, then efficiency is improved for that layout, but adaptability to different warehouse layouts decreases
Solution Approach 1:
The patent implements a universal batching and optimization framework that can adapt to different warehouse topologies. The distance-based batching mechanism and route optimization algorithms are designed to work with various warehouse layouts (single-aisle, multi-aisle, grid-based, random storage) by calculating distances based on the specific warehouse geometry provided. This makes the system universally applicable across different warehouse configurations while maintaining picking efficiency.
Solution Approach 2:
The patent employs dynamic batching and routing that adapts to the specific warehouse topology provided. The system calculates distances and groups orders into batches based on the actual warehouse layout, allowing the optimization algorithm to dynamically adjust to different topologies rather than being fixed to a specific layout type.
Data Source
Figure 1A~1B
Figure 1C
Figure 2
AI summary
A computer-implemented method for determining a picking route for order items, comprising receiving a plurality of orders, each order comprising one or more order items, generating a network representing the locations of order items by generating a plurality of nodes representing locations of order items, each node corresponding to a shelf, and generating directional links between each node, determining a distance measure between each pair of orders, determining one or more batches based on the determined distance measure, each batch comprising two or more orders, and determining a picking route for each batch based on parameters of the order items in the batch.