by 
AR isn’t one technology but a convergence of several that came before it. They include everything from computer vision to AI to 3D modeling. And just as it borrows from other tech, it also gives back. In AR’s current era, it’s being applied to everything from gaming to productivity.
One of those areas that continues to develop is brick & mortar commerce. This extends from the benefits that AR has demonstrated in eCommerce, such as virtual try-ons. In physical retail, AR materializes in related but different ways — both business-facing and user-facing.
Business-facing applications include digital twins of store interiors that help retailers optimize spaces for optimal foot traffic and product placement. And customer-facing applications include in-store wayfinding and informational overlays on products and store shelves.
We examined such possibilities in Part 1 of this series (Augmodo) and Part 2 (Auki Labs). These involve a combination of computer vision and AI to scan store aisles and interiors. This unlocks spatial maps and digital twins to optimize things like inventory management and merchandising.
Practical Result
Building on those narratives, the latest example to cross our desks is Starbucks. Though more in the QSR vertical than pure retail, it’s applying approaches similar to the above at the 3-way intersection of XR, AI, and brick & mortar commerce. It’s all about intelligent operations.
Specifically, Starbucks has launched a tool to help store managers automate inventory management using a combination of computer vision and AI. Known as “AI-powered automated counting,” it replaces manual inventory tracking with a scanning process (see video below).
The software resides on store-supplied iPads, on which users scan shelves or fridges to log items in view. Advantages include time savings (8x faster than manual methods) and lessening cognitive load. The practical result is that associates can divert time and bandwidth to other tasks.
Beyond operational efficiencies, this automated and digitized approach is more effective as it cuts human error out of the loop, such as faulty item counts. Altogether, the result is tighter inventory systems, healthier cash flow (via just-in-time inventory), and customer service.
That last part translates to fewer situations where something is out of stock. We’re talking everything from packaged items to drink ingredients. Given a faster and easier way to take inventory, it means that data collection – and thus restocking – is more frequent and up-to-date.
Broader Canvas
Panning back, Starbucks’ integration aligns with an underrated segment of AI that we’ve been tracking. Though generative AI and LLM-based chat engines get most of the attention and excitement, a potentially-larger opportunity resides in the broader canvas of the physical world.
In other words, just like we examined in Part 2 with Auki Labs, most AI that we discuss and use is confined to the web. That’s of course valuable in several ways, but relatively narrow considering the volume of spending that happens offline. It outnumbers eCommerce 9 to 1.
Beyond AI’s potential in the physical realm, Starbucks demonstrates another point of value: saving businesses time. Deviating from AI’s wow factor in tech circles – including generative AI for knowledge workers – we’re talking here about tangible value for Main Street businesses.
The kicker is that this isn’t anything new. Though AI is buzzy, its most valuable applications tap into fundamentals: saving time and headaches. And the active ingredient in that formula isn’t necessarily AI, but another “A” word: automation. We’re now just getting better versions of it.
Back to Starbucks, the new inventory scanning function has already rolled out across North America locations, installed on tablets already held by those stores. That deployment plan raises another smart aspect of the program – a streamlined rollout involving bits rather than atoms.
Header image credit: kevs on Unsplash
More from AR Insider…
