Warehouse Picking Sequence Optimization via Distance Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In current online concierge systems, shoppers face challenges in efficiently finding and picking items from a warehouse, especially when dealing with long item lists and unfamiliar warehouse layouts, leading to unnecessary time spent moving between aisles.
Innovation Solution
An online concierge system generates a suggested picking sequence by calculating pairwise distances between aisles, creating a distance graph, solving the traveling salesperson problem for historical orders, and applying a ranked global sequence to new delivery orders to optimize the picking route.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If items are organized by category or aisle number in the item list, then the item list structure becomes easier to navigate, but shoppers still spend unnecessary time moving back and forth between aisles
Solution Approach 1:
The system performs preliminary optimization by calculating the optimal aisle traversal sequence before the shopper begins picking items. Using historical fulfillment data and warehouse layout information, the system pre-determines the most efficient path through aisles, allowing the shopper to follow an optimized route rather than moving randomly or by aisle number sequence.
Solution Approach 2:
The system changes the ordering parameter of the item list from simple aisle number or category sequence to an optimized sequence based on spatial relationships and historical picking patterns. This reparameterization of the item list transforms it from a static categorical organization to a dynamic spatial optimization that minimizes movement distance and time.
2Productivity
If shoppers are sent to fulfill long item lists in warehouses with unfamiliar layouts, then order fulfillment capacity is maintained, but the time required to complete orders increases significantly
Solution Approach 1:
The system incorporates feedback from historical order fulfillment data to continuously improve picking route optimization. By analyzing actual picking times, movement patterns, and aisle traversal durations from past orders, the system refines its optimization algorithms to better predict and minimize future picking times, creating a self-improving system that adapts to warehouse-specific patterns.
Solution Approach 2:
The system replaces the mechanical reliance on shopper familiarity with warehouse layouts by implementing an information-based optimization system. Instead of requiring shoppers to learn and remember warehouse configurations, the system uses computational algorithms to generate optimized picking sequences, substituting human spatial memory with automated route optimization.
3Device complexity
If shoppers navigate warehouses without optimized routing, then system complexity remains low, but efficiency and productivity are reduced
Solution Approach 1:
The optimization problem is segmented into manageable components: warehouse layout representation, historical data processing, distance graph construction, and sequence optimization. By breaking down the complex routing problem into these discrete segments, the system can process each component independently and combine them to generate optimized picking sequences without requiring overly complex integrated systems.
Data Source
AI summary
An online system receives an order containing a list of items from a user's client device and tracks the current locations of a client device of a shopper within a warehouse. The system applies a trained item sequence model to generate a suggested picking sequence, minimizing time required for the shopper to pick the items. The item sequence model is trained using historical order data, including durations between picking items from different aisles and pairwise distances between aisle locations. The system transmits the suggested picking sequence to the shopper's client device for display. Responsive to determining that the client device of the shopper's location deviates from the suggested sequence, the system dynamically updates the sequence by applying the model to remain items and the shopper's current location.