At the beginning of July 2026, I spotted Simbe’s Tally inventory monitoring robot operating in a Tesco store in the UK.  

This marked the first time that Tesco is trialling this innovative technology in-store, and only the second time we've seen a major UK retailer trialling autonomous shelf-scanning technology in the UK. 

In other markers, Tally has been operating in stores for years. In August 2026, Kroger reached a milestone by deploying its 500^th Simbe Tally inventory robot at a Smith's Marketplace location. 

We decided to sit down with Seamus McHugh, Head of European Market Development at Simbe to understand the solution in more detail and hear about the opportunities and benefits for retailers, suppliers and shoppers. 

Why are retailers investing in robotics and computer vision? 

Retailers are investing because the physical store remains one of the least visible parts of the enterprise, even though nearly every commercial outcome, from product availability to accurate pricing and omnichannel fulfilment, depends on what happens at the shelf. 

Coresight Research found that in-store inefficiencies now cost retailers 6.4% of gross sales annually, while 60% of retailers have scaled or are actively scaling store-intelligence technologies.   

Robotics and computer vision give store teams a continuous, objective view of shelf conditions that manual audits cannot provide, freeing them to serve shoppers instead of searching for problems.  

This is why leading retailers increasingly view shelf digitisation not as an isolated automation project, but as foundational infrastructure for the connected store. 

Which use cases generate the greatest value? 

The greatest value begins with the fundamentals shoppers notice most: whether a product is available, correctly priced and easy to find. These use cases also strengthen ecommerce fulfilment, promotion execution and inventory accuracy by creating a dependable source of shelf-level ground truth.  

At a leading European food retailer, store teams reduced overall out-of-stocks by 25%, improved controllable out-of-stocks by 30% and improved pricing accuracy by more than 70% in paper-tag environments.   

The strongest retailers then build on that foundation, using the same intelligence to improve ordering, merchandising, supplier collaboration and store productivity. 

What results have surprised you most? 

What often surprises retailers most isn't the volume of issues technology detects, but how many are immediately addressable by store teams. Across Simbe’s platform, 43% of the hundreds of millions of out-of-stock instances detected have been controllable, meaning the product was already somewhere in the store but had not reached the shelf.   

A leading European grocer found that approximately 30% of detected shelf gaps could be resolved within 12 hours once teams had clear, prioritised actions.  

The lesson is encouraging: many availability challenges are not immovable supply-chain problems; they are execution opportunities that store teams can solve when given timely, trusted information. 

What has the shopper response been to the robots in store? 

The response has been highly positive when robots are designed around people rather than the other way around. At SPAR Austria, shoppers and associates described Tally as friendly, quiet and safe, and the robot operated throughout the day without disrupting normal store activity.  

An independent study of 400 shoppers also found that Tally positively influenced retailer perception for 65% of respondents, underscoring that thoughtful design can strengthen rather than diminish the human experience of the store.  

Shoppers ultimately care less about the technology itself than the outcome: fuller shelves, accurate prices and store teams who have more time to help them. 

Which KPIs improve fastest after deployment? 

On-shelf availability, controllable out-of-stocks, pricing accuracy and issue-resolution speed tend to improve first because store teams can act on those opportunities immediately.  

At a leading European food retailer, out-of-stocks fell by as much as 25% within the first 60 days, while pricing errors in paper-tag environments improved by more than 70%.  

HomeBase USA achieved similarly rapid gains, for example, reduced pricing errors by 92% and controllable out-of-stocks by 58% before expanding chainwide.  

Over time, those operational improvements compound into stronger sales, inventory productivity, shopper trust and labour efficiency. 

How accurate is the technology today? 

Tally detects 10X more out-of-stocks than manual audits and is 23% more accurate than fixed camera-only solutions and 15% more accurate than mobile phone-only solutions.  

The technology supports daily enterprise operations. Simbe’s computer-vision platform delivers 99%+ SKU-level and shelf label identification accuracy, while continuously adapting to new products, packaging and store environments.   

Equally important are coverage and frequency: a highly accurate snapshot has limited value if it only captures a fraction of the store or becomes outdated before teams can act. 

The industry is therefore moving toward multimodal approaches that combine autonomous robotics, fixed sensing, RFID and AI to match the right technology to each part of the store. 

What operational changes are required for success? 

The most successful retailers treat shelf intelligence as a new operating capability, not a technology installation. They establish clear ownership, integrate prioritised actions into existing store routines and involve store leadership early so associates understand both how to use the information and why it matters.  

At the leading European grocer, local-language training, regular feedback loops and iterative refinement helped teams incorporate shelf data into daily execution without disrupting operations.  

The technology supplies visibility, but retailers create the value by empowering their people to act consistently and quickly. 

What hidden store issues does the technology uncover? 

Continuous shelf visibility frequently exposes phantom inventory, misplaced products, incorrect or unlinked price labels, promotion errors and stock that reached the store but not the shelf.  

These issues often remain invisible in enterprise systems because the system of record may say an item is available even when the shopper cannot find it.  

At a leading European retailer, autonomous scanning uncovered both paper-price errors and unlinked electronic shelf labels, reducing the latter by 45% and demonstrating that even digitised pricing systems benefit from independent shelf validation.  

This gives retailers a clearer distinction between supply constraints and execution gaps, so  store teams can focuswhere they can have the greatest impact. 

How is AI moving retailers from insight to action? 

The important evolution in retail AI is a shift from describing what happened to prescribing what should happen next.  

Instead of asking store teams to interpret another dashboard, AI can prioritise the highest-impact availability, pricing or merchandising actions by location and direct them to the precise shelf that needs attention.  

Simbe’s platform, for example, analyses visual information against a database of more than 43 million unique product SKUs, price tags and promotions to identify out-of-stocks, misplaced products and pricing discrepancies in real time.   

This closes the gap between detection and execution, making AI useful to the people running stores - not only to analysts at headquarters. 

What opportunities does this create for suppliers? 

For suppliers, accurate shelf-level data creates the opportunity to collaborate with retailers around a shared version of reality.  

Brands can better understand whether lost sales are being driven by supply constraints, replenishment execution, pricing errors, promotion compliance or product placement, rather than relying primarily on shipments and point-of-sale data.  

This enables more productive conversations about availability, merchandising and demand, and helps store teams resolve issues faster.  

The most valuable model will remain retailer-led: retailers should govern how the intelligence is used, while suppliers contribute resources and expertise that improve execution for the shopper. 

What should grocery leaders do now to prepare for the connected store? 

Connected stores are built on intelligent, self-optimising foundations that unify physical and digital channels across all channels.  

Grocery leaders should begin by establishing a trusted, continuously updated view of the shelf, because every downstream system, from forecasting and ecommerce to workforce management and supplier collaboration, is only as effective as the underlying store data.  

Coresight Research’s latest work argues that technology sequencing is a critical determinant of value and that shelf digitisation should precede systems that depend on accurate shelf-level inputs.  

Leaders should also define the operating model early: who acts on insights, how success is measured and how associates will be involved in designing the workflow.  

SPAR Austria’s expansion following a successful live-store trial illustrates the right approach - prove value in real operations, listen closely to store teams and scale from a foundation of shopper and associate benefit.  

Need more insights on the digitalisation of stores  

To help industry understand the complexity, the direction of travel, and current case studies, we have created the following thought leadership reports and articles:  

• Report: The hyper-connected store -  subscribers only.  

• Report: Technology to digitalise the store - subscribers only.  

• Article: Dawn of the hyper-connected store  – free to access.  

• Article: How do you digitalise physical stores? – free to access. 

Subscribe to IGD for in-depth insights and reports on the grocery market  

Share your story with the food & grocery industry  

We welcome contributions from retailers, brands, and solution providers shaping the future of food and grocery. If you have a compelling case study or perspective on technology, operations, or shopper behaviour, we’d be keen to hear from you.  

Please get in touch with [email protected]