Findability Machine-Learning Model for Item Placement Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Retailers face challenges in accurately determining the ease or difficulty for customers to find items within their physical stores, as existing methods like questionnaires are subjective and time-consuming, lacking an effective measure for optimizing item placement.
Innovation Solution
An online concierge system employs a machine-learning findability model to predict item findability by training on data from pickers, using item maps and calculating findability scores based on average time and success rate, iteratively optimizing item placement to improve findability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If questionnaires are used to measure item findability, then customer feedback can be collected, but the measurement is inaccurate and time-consuming
Solution Approach 1:
The patent replaces the mechanical questionnaire system with an automated machine learning model that computes findability scores based on picker behavior data. The findability model processes objective metrics (time to find item, success rate) automatically, eliminating the need for manual questionnaire administration while providing more precise measurements.
Solution Approach 2:
The system enables self-service measurement by having pickers naturally interact with the item locations during their regular work, and the system automatically tracks their behavior. This eliminates the need for separate data collection processes, as the findability data is gathered passively through normal operations.
2Reliability
If traditional item placement methods are used, then item organization is simple, but item findability is poor
Solution Approach 1:
The system performs preliminary actions by pre-training the findability model on historical picker data and pre-computing findability scores for different item map configurations. This allows the system to evaluate multiple placement scenarios before implementing changes, ensuring reliable findability improvements without ad-hoc trial and error.
Solution Approach 2:
The patent implements a feedback loop where findability scores computed by the machine learning model are used to evaluate item map configurations and guide optimization. The system continuously refines item placement based on feedback from the model, improving findability while managing complexity through automated iteration.
Data Source
AI summary
An online concierge system uses a findability machine-learning model to predict the findability of items within a physical area. The findability model is a machine-learning model that is trained to compute findability scores, which are scores that represent the ease or difficulty of finding items within a physical area. The findability model computes findability scores for items based on an item map describing the locations of items within a physical area. The findability model is trained based on data describing pickers that collect items to service orders for the online concierge system. The online concierge system aggregates this information across a set of pickers to generate training examples to train the findability model. These training examples include item data for an item, an item map data describing an item map for the physical area, and a label that indicates a findability score for that item/item map pair.