The Power of Sensors: How RADAR is Simplifying Customer Experience Operations for American Eagle

Finding the right item is a basic customer-service task that becomes difficult when store inventory is unreliable. In this Retail Refined interview from Shoptalk 2023, Melissa Gonzalez talks with RADAR CEO Spencer Hewitt about ceiling-mounted sensors and their role in American Eagle’s store operations.

Hewitt describes a system combining RFID and computer vision to locate products and support associates. The conversation centers on a pilot, an announced expansion to roughly 500 American Eagle stores and the accuracy he says supported that decision. It also separates that inventory work from the analytics and autonomous-checkout products he described as coming next.

What RADAR’s ceiling sensors are designed to measure—and why that matters

Hewitt frames RADAR as a sensing system installed on the ceiling that can locate “every item and every person” in 3D by combining RFID with computer vision. In his telling, that fusion is what enables not just counts, but spatial specificity—knowing what is in the building and where it physically sits, rather than inferring inventory from sales and periodic cycle counts.

He ties that measurement directly to common store-floor friction: associates spend time searching for items, verifying whether something is actually in stock, and fulfilling online orders from the store. Hewitt says real-time inventory accuracy is intended to shorten those searches, reduce “is it here?” uncertainty, and help teams find product for both walk-in customers and order pickers.

American Eagle’s rollout: how a pilot became “ready for fleet”

A key episode moment is the ShopTalk announcement: Hewitt says American Eagle Outfitters—RADAR’s first customer—planned a rollout to about 500 stores. Gonzalez and Hewitt frame the goal of the partnership around customer experience, with “inventory accuracy” as the operational lever: the right products, in the right place, at the right time across the chain.

When Gonzalez asks what success looked like in the pilot phase, Hewitt keeps it narrow and measurable. He says the pilot focused on a handful of stores and tested whether RADAR could maintain consistently high accuracy when benchmarked against manual audits—essentially, whether the system could replicate the confidence of continual auditing, but done automatically and repeatedly.

Hewitt reports that the pilot achieved about 99% accuracy against manual audits, which he says helped American Eagle decide to expand. The figure is his account of the pilot’s performance. In his description, consistently accurate inventory was the immediate priority before introducing additional products and use cases.

From inventory to analytics, monetization ideas, and autonomous checkout

Hewitt positions analytics as RADAR’s next product direction after establishing inventory accuracy. He contrasts this with simpler store reporting, suggesting the goal is to understand what shoppers do with specific items, not only where people congregate. In his outline, the point is to extract higher-resolution behavioral signals—what gets handled, rejected, and ultimately purchased—using the same sensing foundation.

He also describes “autonomous checkout” as a later step, enabled by knowing the position of every person and every item. Hewitt explains the mechanism at a high level: if the system can attribute products to individuals as they leave the store, it can support automatic charging. Gonzalez connects that to broader industry interest in walk-out-style experiences, while Hewitt presents it as an extension of accurate attribution.

Looking further out, Hewitt sketches an additional implication for large multi-vendor retailers: the potential to monetize in-store browsing data with vendors and connect it to their e-commerce ad networks. He argues that retailers already invest in digital ad networks, but lack cost-effective in-store measurement at comparable scale. In his scenario, in-store browsing signals could be packaged for vendors and tied back to online ads that convert better, creating multiple revenue touchpoints.

What “AI” means here—and the associate impact RADAR didn’t expect

When Gonzalez asks about AI’s role as deployments grow, Hewitt describes specific uses rather than a general label: deep learning for computer vision to extract people’s 3D pose in real time, and machine learning in RFID location algorithms. He adds that AI could later help generate insights from the broader “walk-out” and analytics technology, but emphasizes that AI already sits in the core functions of the system as he defines them.

The most surprising learning Hewitt shares is about human adoption, not dashboards. He says store associates responded positively to the apps RADAR provided, and that the tools made inexperienced hires useful to customers quickly. As he tells it, a new worker could learn to locate a product in about five minutes, reducing the need to memorize the store and stockroom. Gonzalez notes that this aligns with ongoing attention on empowering associates with better tools.