The Rise of Personal Trackers as the Next Infrastructure for Ubiquitous Positioning
Reliable indoor positioning remains a central unsolved problem in spatial computing. Existing systems depend on fixed, provider-owned infrastructure (WiFi access points, Bluetooth beacons, UWB anchors) whose deployment, calibration, and ownership models limit their scalability. Meanwhile, personal trackers, wearables, smart luggage tags, and tagged industrial assets are becoming densely embedded in the very indoor environments where positioning infrastructure is weakest. This paper argues that these devices point to a new class of spatial infrastructure: non-stationary, user-owned, and semantically interpretable. We present TrackerNet, a vision for transforming personal trackers from private object-finders into a decentralized substrate for ubiquitous positioning. The key shift is to stop treating anchor mobility as noise: TrackerNet models mobile anchors as probabilistic references whose stability, trust, and semantic role are inferred over time, reframing indoor localization as non-stationary spatial graph inference. We identify three research frontiers (non-stationary SLAM over mobile, intermittently visible anchors; semantic reasoning that connects ranging signals to interpretations of anchor behavior; and consent-based, privacy-preserving governance for cross-vendor cooperation) and a staged pathway by which the vendors and venue operators who already run tracker networks can realize them with academia. We do not claim a finished system; we define a research agenda.