RussiaRetail and e-commerce2021 – ongoingMagnit
Neural-network shelf recognition for planogram checks
Magnit cut shelf-check time from about an hour to 15 minutes per employee with photo recognition
Problem
Staff checked by hand whether products sat on shelves as the planogram required. Magnit described the manual method as not accurate enough.
Approach
Magnit piloted a neural network in 20 stores from September 2021. It recognizes products in shelf photos, compares them with the planogram and sends corrections to staff through a mobile app. In November 2022 the chain began rolling the system out to about 1,000 stores across its main formats, with a decision on wider scaling due after the second quarter of 2023.
Results
Takeaway
The pilot results came from the chain's own stores and were stated by a Magnit executive. The wider rollout decision was still open when the news was published.
Categories
- Size
- Large enterprise
- Industry
- Retail and e-commerce
- Business function
- Production and quality
- Innovation type
- Process innovation
- AI technology
- Computer vision
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
Retail.ru reports Magnit's statement on its shelf recognition pilot in 20 stores and the planned rollout to about 1,000 stores. It gives time saved, a sales uplift and an accuracy figure, all from Magnit's deputy general director.
Full details are in the original sources.
