500 Robots - 220m Records AI Goes Live

This week, 500 robots went live inside Kroger's stores, Delhaize hit 90% forecast accuracy with AI, and McDonald's built a system trained on 220 million customer loyalty records. So today, we're asking, what happens when AI stops being a pilot and starts being infrastructure? Hi, I'm Clare, and welcome to this week's Retail Reality Check, a podcast on the latest tech installs in retail and consumer product goods, brought to you by IHL Group. Before we dive into today's stories, a quick invite. IHL, along with Connexion and Dell, is hosting a live webinar called AI Reality Check: What Retailers Need to Get Right Before Scaling AI on August 26th at 2 PM Eastern. We'll be talking about what AI readiness actually looks like at the store level, and why the next wave of retail transformation is going to be driven by investment in infrastructure, data, and edge technology. Registration is in the episode description. We'd love to have you there. To set the stage, here's Greg Buzek with this week's take.

Claire, this week's stories all point to the same shift. AI-driven replenishment has moved past the pilot stage into daily operations. And robotics has crossed from experiment into infrastructure. Inventory distortion is not just a forecasting problem upstream. It is also a shelf-level visibility problem downstream, and the retailers making real progress are the ones connecting both ends. The blueprint is simple to describe and hard to execute: integrate forecasting with the system of record, measure accuracy relentlessly, and let automation absorb the routine work.

Thanks, Greg. Let's get into it. Let's start with Delhaize. The Belgian grocery chain, part of Ahold Delhaize, rolled out Symphony AI's demand forecasting and replenishment platform across 600 affiliated stores and 5 distribution centers, integrated directly with SAP Retail. Weekly forecast accuracy at the distribution center level is now running 80 to 90%, And store-level accuracy is up as much as 10 points in seasonal locations, while shrink and safety stock have both come down. IHL's 2026 Inventory Distortion Study backs this up directly. Demand planning and forecasting delivers positive results for 76% of AI and machine learning deployers, the highest success rate of any application IHL measured. Delhays' numbers track that pattern closely. Next, Salengroup, Northern Europe's largest retail group. They just modernized their ERP to SAP S/4HANA Cloud through RISE with SAP, which is less an AI story and more a foundation story, laying the cloud groundwork before layering AI on top. They're also running SAP Extended Warehouse Management across 29 distribution centers. Salling's SAP-first approach is exactly the kind of unified foundation that data says AI investment needs to pay off. Then there's Simba. 500 of its tally inventory robots are now live across Kroger's store network, with the milestone hit at a Smith's Marketplace location. These autonomous robots use computer vision to scan shelves and flag out-of-stocks, price errors, and planogram gaps without pulling associates off the floor. IHL's research on RFID and computer vision documents a comparable case: a major sporting goods retailer using an autonomous RFID-reading store intelligence platform to speed up replenishment accuracy. Retailers combining robotics with store intelligence report out-of-stock reductions near 60% and on-shelf availability near 98%. Now, McDonald's. They're consolidating loyalty data from 220 million MyMcDonald's Rewards users across 70 markets into a single global AI data lake built on a Google Cloud-powered restaurant platform edge that already processes about 120 signals per order in real time. IHL research found retailers using AI and machine learning in order management expect 92% higher profit growth than non-users, with the top performers achieving positive AI and ML results at twice the rate of laggards. McDonald's is betting its loyalty data platform delivers similar returns, just at a scale nobody else in retail is playing at. And Albertsons— they just launched a Safeway plugin inside ChatGPT, letting shoppers reorder favorites, shop from a photo or a recipe, and build a cart in natural language before checking out on Safeway's own platform. IHL's AI Resilience Research found that customer-facing AI tools like chatbots are becoming table stakes, with a pretty limited competitive moat. While the durable advantage comes from back-office integration. Albertsons' plugin is a smart move, but it needs to be paired with deeper inventory and fulfillment AI investment to hold its edge over time. That's the thread running through our deep dive stories this week. The technology to fix inventory distortion already exists. The gap isn't invention anymore, it's adoption. Connecting the forecasting, the robotics, and the data in into one system instead of 5 separate pilots. Now, in our lighting round, let's move fast through the rest of this week's news. Walmart is deploying 90 million Wiliot IoT tags by the end of 2026. The battery-free IoT pixels track temperature, humidity, and location in real time to protect perishables and verify cold chain compliance. Boot Barn's e-commerce sales grew 13.4% in its fiscal first quarter, helped by BOPIS, ship-to-store, TikTok Shop, and Salesforce's Agentic AI tools for product discovery. Morrisons is rolling out Everseen's EverCheck computer vision platform across 200 UK stores after a successful trial, using on-screen prompts to catch missed items at self-checkout. Hy-Vee has deployed Nations Benefits Basket Analyzer chain-wide, letting health plan members buy eligible food, wellness, and over-the-counter products at checkout without filing separate claims. Furniture.com partnered with FlexPay to launch a single checkout financing experience that lets shoppers furnish a room across multiple retailers with one application and one monthly payment. Academy Sports and Outdoors is now offering 1-hour delivery through Instacart across more than 300 stores in 21 states, its 3rd on-demand delivery partnership. Amazon expanded its Locker network to more than 750 pickup points across 500 college campuses ahead of back-to-school season. Bush's Fresh Food Market is rolling out Flash Foods' Surplus Grocery app across all 16 of its Michigan stores to cut food waste on discounted produce, meat, and dairy. BWH Hotels, which includes Best Western and World Hotels, endorsed Toast as its point-of-sale provider for US and Canada properties. Kroc signed a 10-year AI transformation agreement with Infosys to unify enterprise data and break down operational silos. CKE Restaurants, the parent of Hardee's and Carl's Jr., consolidated 22 point-of-sale platforms and 14 menu configurations into one centralized database, Cutting product launch time from 2 weeks to 2 days. Schreiber Foods is partnering with Ascendian on a multi-year deal to bring Agentic AI into ERP modernization and food safety monitoring across its global operations. And Wayfair says its luxury banner Perigold is generating more than $400 million a year with an in-house AI pipeline that cut room imagery production costs by more than 99%. So here's the question worth sitting with this week: if the technology to fix inventory distortion already exists, what's actually stopping your organization from deploying it? Is it the technology? Your data? Your infrastructure? Or is it everything around the technology? That's it for this week. Head to ihlservices.com for more than 50 additional stories we didn't have time to cover here. Plus, free white papers on everything from inventory distortion to computer vision in retail. Thanks for listening, and we'll see you back here next week.