Note that these are all significant improvements on the more vanilla versions of these use cases: (a) Reminders based purely on in-store location; (b) a shopping list that you must manually track on your smartphone, and (c) recommendations based purely on general, longer-term customer profiles. Our vision engenders the following key question: “How can applications use infrastructure sensors, and wearable/mobile devices to unobtrusively obtain deeper insights on customer interactions?” We propose an architecture where low-cost BLE beacons+ embedded sensors are mounted on product shelves, and their data is fused with sensor readings from a smartwatch worn by a shopper.