AI ambassadors that show you around the store. Smart shelves that never go bare. Checkout that just happens without any scanning. Robots moving inventory. Advertising that changes in real time according to your current needs and interests.
This sounds like something out of a sci-fi movie, but we’ve seen some of these technologies exhibited recently. From CES to NRF’s Retail Big Show, 2026 is shaping up as a pivotal year for Artificial Intelligence (AI) in retail. CES showcased a surge of robotics and embodied AI, while NRF highlighted accelerating investment in agentic and phygital retail technologies. The excitement is real, but the question remains: can these systems deliver measurable commercial impact beyond controlled trials?
The short answer – it depends on infrastructure.
Physical AI as the new retail differentiator: from pilots to production scale
Pilot programs have demonstrated physical AI’s tremendous potential. IoT sensor implementation has steadily built synergy between the physical and digital in retail, paving the way for AI integration. Though the technology still has some maturing to do, its real-world applications are undeniable – and early adopters are already separating from the pack.
The business case is compelling. Physical AI can increase basket sizes, eliminate queues, keep shelves stocked and make the workforce dramatically more efficient. For retailers navigating razor-thin margins, this isn’t experimentation – it’s a competitive imperative.
Intelligent checkout systems maintain up-to-date inventory tallies and manage customer flow, helping to prevent bottlenecks. Cameras with computer vision track products as customers add them to baskets, then AI processes transactions at the point of exit – payment without traditional scanning. This seamless experience can increase basket sizes whilst reducing operational costs and shrinkage.
AI-enhanced workforce tools make staff significantly more efficient. Armed with real-time product information and wayfinding capabilities, they can answer customer questions and locate products instantly – no more wild-goose chases. Smart shelves can proactively alert staff when inventory runs low and suggest optimal replenishment timing based on foot traffic patterns.
Remarkably, over two-thirds of consumers have left stores without purchasing anything simply because they couldn’t find what they needed. Physical AI is designed to eliminate this friction entirely. Tracking customer movement allows retailers to enhance the in-store experience dynamically – analysing which displays attract attention, how long customers linger, where bottlenecks occur. The store evolves from a static environment to one that continuously learns. No more assumptions. Just behaviour-driven intelligence.
Turning aisles into experiences: the rise of the Phygital Brand Ambassador
The solutions above enhance what happens behind the scenes. But physical AI’s real power emerges when it transforms the customer-facing experience itself.
E-commerce was supposed to kill brick-and-mortar. It didn’t. Physical stores still account for nearly 80% of retail sales. The opportunity is clear: combine the immediacy of being in-store with intelligent, adaptive environments that deliver moments online simply cannot match.
This marks the beginning of a retail renaissance – physical stores reimagined as immersive, intelligent destinations. At the centre of this shift is the Phygital Brand Ambassador: a fusion of AI, adaptive intelligence and near real-time connectivity that redefines customer engagement. These aren’t simple chatbots. They’re AI embedded into the store environment itself – virtual representatives explaining brand heritage, AR athletes demonstrating products in-aisle, entertainment-led commerce that transforms routine shopping into memorable moments.
The breakthrough lies in seamless human-AI handoffs. Virtual ambassadors handle initial engagement and routine enquiries, analyse customer preferences and purchase history, then transfer to human associates for high-value relationship moments. Staff are freed from time-consuming wayfinding to focus on what actually drives loyalty: genuine human connection.
But here’s the catch: when customers interact with AI-powered displays, AR wayfinding systems or virtual brand representatives, even a delay of a few milliseconds can potentially disrupt the experience. Connectivity isn’t a nice-to-have – it’s the defining factor.
No infrastructure, no intelligence: why connectivity is non-negotiable
Physical AI in retail can be deployed today. But here’s the caveat that trips up most retailers: it demands the right infrastructure. The technology is proven – from computer vision to autonomous robots – but these applications require high speeds, low latency and substantial bandwidth. Most store networks simply aren’t built for this.
Large retailers now manage a vast ecosystem of IoT devices per store – from electronic shelf labels and RFID readers to computer‑vision cameras, environmental sensors and point‑of‑sale systems. Each relies on real‑time data processing. Without purpose‑built connectivity, the result is a bottleneck; with it, retailers unlock actionable intelligence.
Edge computing processes information where it matters – on the shelf, at checkout, on the shop floor. Smart shelves detect out-of-stock conditions in near real time, triggering automatic replenishment before customers see empty displays.
Edge and private 5G form the backbone. Edge computing nodes process AI inference locally, reducing latency and lowering bandwidth costs whilst keeping sensitive data on-premises. Private 5G provides deterministic, reliable connectivity – because when autonomous checkout processes transactions, “usually works fine” is not good enough. Physical AI demands consistent performance.
Investment, not expense: the strategic infrastructure imperative
Retailers investing in sophisticated AI tools must also invest in connectivity infrastructure. Those pioneering physical AI transformation share a common characteristic: they’ve elevated networks from cost centres to strategic enablers, recognising connectivity as the foundation upon which everything else is built.
The financial outlay may give some pause. But weigh that investment against what you stand to lose. Brick-and-mortar retail remains the biggest revenue generator in its industry. Failing to strategically enhance the in-store experience isn’t caution – it can be a missed opportunity at scale.
The bottom line
Physical AI is a watershed retail innovation. Online shopping may have squeezed margins, but it has also laid the groundwork for a new retail paradigm – one that seamlessly blends digital capabilities with physical spaces. The future belongs to retailers willing to move now. Everyone else will be playing catch-up.
