While the retail world obsesses over generative AI and chatbots, a quieter revolution is happening on store shelves and warehouse floors. It’s called physical AI, and it represents the convergence of two powerful forces: ambient IoT sensors that can track anything, anywhere, and artificial intelligence that can turn that data into instant operational decisions.
Doron Hazan, Director of Data, Product, and AI at Wiliot, recently joined OmniTalk to explain how this technology is already transforming retail operations at scale, as evidenced by a massive deployment across Walmart stores. The implications go far beyond simple inventory tracking. We’re talking about retail operations that can sense, learn, and respond in real-time.
What Makes Physical AI Different From Digital AI
Most AI applications today operate in the digital realm. They’re trained on text, documents, spreadsheets, and internet data. Physical AI takes a fundamentally different approach by grounding artificial intelligence in the actual, physical world.
“Digital AI is what we’re familiar with,” Hazan explains. “It’s being trained on the internet, on documents, text, spreadsheets. But it’s not really connected into the physical world. Physical AI is being trained on physical things—where they are and their condition.”
The breakthrough comes from combining ambient IoT with AI capabilities. Wiliot’s approach uses battery-free Bluetooth sensors called “IoT pixels” that attach to products, cases, and pallets. These tiny sensors continuously capture data about location, temperature, humidity, movement, and dwell time, all without batteries, scanning, or human intervention.
That ambient data flows into Wiliot’s intelligence platform where machine learning models transform raw sensor information into actionable insights: real-time inventory accuracy, freshness detection, shrink prevention, and automated replenishment triggers.
The Customer Experience Improvements You’ll Never Notice
Here’s the paradox of great physical AI: when it works, you don’t see it. The technology operates invisibly in the background, but the benefits land directly with customers.
Hazan outlined three major customer-facing improvements. First, shoppers encounter fewer out-of-stocks because retailers know exactly what needs replenishing and when. Second, food stays fresher because temperature excursions get flagged immediately, even during transport. Hazan shared a compelling example about strawberries that get accidentally frozen in truck AC systems. Without continuous monitoring, those strawberries might look fine when they arrive but are already damaged. Physical AI catches these issues before customers ever see compromised products.
Third, click-and-collect and delivery experiences improve dramatically. When retailers have real-time visibility into exactly where items are and their condition, they can assemble orders more accurately with fewer errors and delays.
The common thread? None of this requires customers to interact with new technology or change their behavior. They simply experience better availability, better freshness, and better fulfillment.
Making Store Operations Smarter, Not Harder
While customers experience the benefits passively, store operators see dramatic workflow improvements. Physical AI doesn’t replace human workers. It empowers them with better information and automated alerts.
“Physical AI doesn’t replace people, it empowers them,” Hazan emphasizes. “The ultimate model is a partnership between physical AI and humans.”
For warehouse and store teams, this means less manual work. No more searching for missing pallets, walking around scanning items, or manually checking temperatures. The system handles routine monitoring automatically while alerting associates to exceptions that need human attention.
Store operators gain what Hazan calls a “live operational picture,” i.e. a real-time view of what just arrived, what needs to be moved, what’s sitting too long, and what’s approaching temperature thresholds. Associates receive guided workflows rather than generic reports they have to interpret and act on later.
The vision Hazan describes sounds like mission control for retail operations, a centralized dashboard showing all inventory exceptions and operational priorities in real-time. It’s the kind of complete visibility that train operators have for their networks, applied to the movement of products through stores and distribution centers.
The Platform Approach: Start Small, Scale Smart
Given the transformative potential of physical AI, retailers might be tempted to deploy it everywhere all at once. Hazan cautions against this approach.
“We should start small,” he advises. “Start by identifying a real problem with real business value, scope it well, and use physical AI to solve that problem. If you solve that problem and you can see it generates ROI and value, then you can scale into different use cases.”
The key is starting small with a platform approach rather than building disconnected point solutions. Wiliot’s intelligence platform handles the “static friction” of initial setup, e.g. infrastructure requirements, integrations, security checks, and system connections. Once that foundation exists, adding new use cases and scaling across locations becomes much easier.
This platform thinking addresses a common frustration Hazan hears from grocery retailers in particular: too many dashboards and point solutions that don’t connect into a unified operational brain for the store.
What’s Next: The NRF 2026 Announcements
Hazan remained deliberately cryptic about specific product announcements, but he confirmed that Wiliot will unveil major updates to both their IoT pixel sensors and intelligence platform at NRF 2026 in January. The improvements span performance, range, accuracy, and AI processing capabilities.
More importantly, Wiliot is focused on making the entire platform easier to deploy, integrate, and scale across large distributed retail networks, all of which is crucial for the kind of enterprise deployments they’re executing with Walmart.
Hazan believes we’re entering a moment where physical AI becomes the new competitive edge in retail. “Not just digital analytics, not just forecasting, but real-time sensory ground truth data,” he explains. “Retailers who adopt this now are building automated workflows and smarter supply chains—almost self-aware operations.”
That vision of self-aware retail operations, where physical infrastructure can sense its own state, identify problems, and guide human responses, represents a fundamental shift in how stores and distribution centers function. And unlike many emerging technologies that promise transformation, physical AI is already operating at scale in thousands of locations.
The technology works best when customers never notice it. But for retailers wrestling with inventory accuracy, freshness management, and operational efficiency, physical AI offers a pathway to operations that finally know what’s happening in real-time.
Listen to the full episode to hear Doron Hazan extol on the wonders of Physical AI wherever you get your podcasts.
Apple Podcasts | Spotify | SoundCloud | Amazon Music
Be careful out there,
– Chris, Anne, and the Omni Talk team
P.S. See our past 8 years of wonderful Spotlight Series podcast guests, featuring roughly 200 movers and shakers in retail, by clicking here
Be careful out there,
– Chris, Anne, and the Omni Talk team
P.S. See our past 8 years of wonderful Spotlight Series podcast guests, featuring roughly 200 movers and shakers in retail, by clicking here
#physicalai #retailtech #ambientiot #wiliot #omnitalk #smartretail #retailinnovation #inventorymanagement #supplychain #retailpodcast
Music by hooksounds.com
*Sponsored Content*
Omni Talk® is the retail blog for retailers, written by retailers. Chris Walton and Anne Mezzenga founded Omni Talk® in 2017 and have quickly turned it into one of the fastest growing blogs in retail.