Skip to main content

Shelf Inspector

AI-driven object recognition to ensure on-shelf availability

Accelerate your retail execution and move from manual checks to real-time shelf intelligence.

Shelf Inspector is a complete end-to-end service designed to tackle the monitoring of product displays by retail partners, managing everything from photo collection to meaningful KPI insights. While traditional shelf display inspections are often pricey and found to be invalid in 15-40% of cases, this computer vision application allows retailers to assess display quality standards quickly and objectively using portable devices. The application utilises a neural network and a model trained by Azure Machine Learning to provide certainty that products remain constantly available on the shelves.

Snimek-obrazovky-2022-01-04-104

By analysing a single photo of a supermarket shelf, the solution identifies product positioning, pricing, and out-of-stock items.

It replaces labour-demanding manual counting, which is extremely prone to human error, with a scalable digital process that integrates real-time data into your business intelligence.

shutterstock_2617361313

For FMCG producers

Ensure your products reach the right customers with a precise and objective shelf monitoring solution that tracks real-time product availability and display quality.

shutterstock_2060828513

For field merchandisers

Spend 80% less time on every store visit by using a mobile app that automates data collection, allowing you to focus on higher-value tasks.

4people_business_meeting_orange

For trade & category managers:

Gain deep insights into shelf share, analyse competing goods, and receive immediate alerts regarding incorrect prices or promotion compliance.

shutterstock_2477764607

For supply chain teams

Integrate real-time stock level data directly into supply chain planning to improve replenishment and reduce category-wide out-of-stocks.

Key features

1
Mobile app for photo collection:
 
Developed in close coordination with field reps to ensure seamless data collection.
2
Computer vision model:
 
Capable of detecting SKUs, recognising prices, and assessing planogram compliance.
3
Modular architecture:
 
A "white-box" approach that allows for the quick addition of new technologies or client-specific metrics.
4
Complex reporting BI app:

Provides management with a comprehensive KPI overview, including out-of-stock trends and price developments.
5
Planogram digitalisation:

Create digital planograms from shelf photos even when official documentation is unavailable.
6
Supply chain integration:
 
Streamline replenishment by leveraging real-time data synergies with retail partners.

How it works

The implementation of Shelf Inspector typically follows a structured 2-3 month pilot project to ensure high delivery confidence. In Month 1, data collection begins using the dedicated mobile app, which is compatible with both Android and iOS. During Month 2, the neural network detection models are adjusted to specifically identify your product categories and price tags60. By Month 3, reporting is customised into Power-BI dashboards, delivering out-of-stock alerts, display timelines, and competitive trend analysis.

growth_icon_outline_dark

End-to-End Security

Full supply chain transparency, advanced hardware protection, and trusted execution technologies.

end_to_end_icon_dark_outline

Energy Efficiency

Reduced power consumption per workload with eco-conscious infrastructure design and smart VM scheduling.

end_to_end_icon_dark_outline

Easy Expansion and Automation

Modular, composable hardware with API-driven orchestration for seamless scalability.

end_to_end_icon_dark_outline

Easy Deployment

An already validated reference architecture and a SPOC for hardware-software integration and support.

The power of Shelf Inspector

objective
~99% recognition accuracy

 Achieve precise product recognition that can distinguish between SKUs even when designs are highly similar. 

stopwatch
80% faster store visits
 Automated checks allow field staff to reduce time spent per visit by 80%, shifting focus to value-added retail activities.
burndown
20% out-of-stock reduction

 Proven results show a significant decrease in out-of-stock incidents across multiple categories for global market leaders. 

icon-black-finance
Dynamic price intelligence

 Centralise the management of prices across different retailers and monitor competitive pricing and promotion detection. 

rapport-de-donnees
Precise shelf share metrics

 Monitor your actual share of the shelf in every category and compare it directly against contracted planograms. 

apps
Real-time quality control

 The accompanying mobile app provides immediate feedback on photo quality, identifying blurry images to ensure high-quality data detection. 

Success stories

left-quote Created with Sketch.

 "With Shelf Inspector we are looking at 20% out of stock reduction across multiple categories. It helps us consistently improve our store layouts and price monitoring." 

Tibor Molnar

National Sales Manager, Nestlé
left-quote Created with Sketch.

"Thanks to Shelf Inspector, our merchandisers can use their store check for value-added tasks. Also, we can monitor prices of competitive products and receive alerts."

 

 

Petra Diamond

Head of Trade Marketing, Red Bull

A smarter and safer way to scale retail AI 

Every organisation has its own constraints, whether regulatory, operational or financial. Shelf Inspector gives you the freedom to innovate within those boundaries, without sacrificing speed, security or control. It provides a "white-box" and modular solution that integrates with your existing data models, ensuring you stay ahead of the curve while maintaining full transparency and auditability of your retail data. 

FAQ

Manually collected data is highly inaccurate, with studies showing up to 50% inaccuracy in some cases, often due to human error during the counting process. 

A typical one-off pilot involves approximately 20 stores, one retailer, and 50 SKUs within a single category over a 2-3 month period. 

Yes. For example, Nestlé successfully manages over 250 SKUs despite redesigning 15% of their portfolio annually.

Yes. It is designed to track product availability on primary shelves as well as secondary placements and promotional displays. 

The app identifies store locations needing service, reduces unnecessary visits, and provides immediate cloud data access for efficient time management. 

Yes, the modular solution provides price tag analysis that can detect promotions and compare your pricing with competitive products. 

The engine uses neural networks and models trained by Azure Machine Learning to detect "objects of interest" with up to 99% accuracy.

The mobile app includes built-in quality control to detect blurry or unusable photos immediately, prompting the user for a better shot. 

Get in touch