ShelfOptix

ShelfOptix — High-resolution ground-truth shelf imagery captured by a portable in-aisle scanning robot, at detail sufficient for machines to read individual price tags, count facings per product, detect voided slots including shelf depth, and verify planogram adjacency - fused with the retailer's planogram, point-of-sale and on-hand inventory records to yield out-of-stock, phantom-inventory, planogram-compliance and promotion-display compliance findings across many stores. Small specialist vendor (about 8 employees, no disclosed funding) sitting on a category-defining asset: robot-grade high-…

ShelfOptix offers (Alternative, Geospatial, Supply Chain) — Very high on the imagery axis: a full-aisle, close-range, high-resolution capture per store per pass, so order of thousands to tens of thousands of images per store per scan across a syndicated multi-store footprint, yielding an image corpus plausibly in the millions of frames with derived per-slot structured records; the company claims ground-truth data at a scale never before possible.

Training and benchmarking retail shelf computer-vision models (SKU detection, OCR of price and shelf tags, planogram compliance classification, void detection) on real in-store imagery rather than rendered or lab-captured data - this is the single hardest retail-vision dataset to obtain; vision-language model grounding for retail agents; autonomous store and scan-with-your-phone systems; out-of-stock and lost-sales economic modeling; shelf-space and share-of-shelf analytics for brands; consumer-goods execution auditing; robotic navigation datasets for in-aisle autonomous operation

Coverage spans US; Application Software, Food Retail; alternative, geospatial, supply_chain; equities.