NearSight vs cool

See the hidden world around you. NearSight listens for the Bluetooth Low Energy (BLE) signals that nearby devices are constantly broadcasting and decodes what they actually are. Earbuds, watches, fitness trackers, beacons, smart-home sensors, item trackers, and more turn from anonymous radio noise into a clear, readable list right on your iPhone. WHAT YOU CAN DO • Scan live. Watch nearby devices appear in real time, sorted by how close they are. • Know what's there. NearSight recognizes a growing catalog of device types and shows make, model, and what each signal means, not just a cryptic address. • Notice the unexpected. Surface unknown or unfamiliar devices around you and decide for yourself what belongs. • Go deeper. Open any device to inspect its raw advertisement, signal strength, and parsed technical details. OPTIONAL AI INSIGHTS Turn on AI to get plain-language explanations of what NearSight is seeing. Bring your own Anthropic or OpenAI API key, or run a compact model entirely on your device. Your choice. AI is completely optional and stays off until you enable it. PRIVATE BY DESIGN Scanning happens on your iPhone. NearSight works fully offline and shows what it sees to no one but you. AI insights and anonymous telemetry are strictly opt-in and off by default. Nothing leaves your device unless you turn it on. No ads, no third-party tracking, no account required. WHAT NEARSIGHT CAN (AND CAN'T) SEE NearSight decodes the Bluetooth signals that devices broadcast but it can't see everything: • Not every device is recognized. New or niche hardware may show up as Unknown until we've researched it and added support. • Some devices are intentionally excluded. We don't surface certain safety-sensitive devices (for example personal panic buttons), because revealing someone's private safety equipment could put them at risk. However, AI Insights may identify such devices. • It only sees Bluetooth. Devices that are powered off, not currently broadcasting, or that use other radios (Wi-Fi-only cameras, wired, or cellular devices) are invisible to NearSight. NearSight is a Bluetooth explorer, not a safety or security device, bug sweeper, or hidden-camera detector. It cannot guarantee it will find every tracker, camera, or listening device near you, and not seeing something here does not mean it isn't there. If you have a real safety concern, use dedicated tools and contact local authorities. Questions, or a device you'd like us to identify? Email hello@nearsight.one.

cool-radar turns the radios already inside your iPhone into a radar for the devices around you — with no extra hardware, and without inventing a single number. WHAT YOU SEE Nearby Bluetooth LE devices appear as contacts with an estimated distance, a bearing once there is enough geometry to support one, and a ranked guess at what each device actually is. The Wi-Fi card shows the link your phone is really on: type, interface, local address and subnet, the default gateway and the gateway's manufacturer. Where both ends have UWB hardware, distance stops being an estimate and becomes a measurement. HONEST BY DESIGN Every value carries a badge. "Measured" is a hardware reading. "Inferred" is a value the math derived. Distances read "≈ 4.5 m" because radio ranging is an estimate, not a ruler. When there is not enough geometry for a direction, the app draws a ring of possible positions and a question mark instead of a convincing-looking arrow. And when iOS simply does not expose something — neighbouring Wi-Fi networks, raw spectrum, Wi-Fi CSI — the app says so instead of filling the gap with fiction. THE MATH INSIDE • Log-distance path-loss inversion, with a per-device exponent that recalibrates itself from the geometry you walk • Kalman filtering with motion-gated noise, so a device that is standing still stops drifting on screen • Robust multilateration with Huber weights, so a single reflection off a wall cannot drag a position across the room • A naive-Bayes hypothesis engine that ranks device types from the broadcast name, company ID, advertised GATT services, transmit power, address policy and advertising cadence — and shows you the evidence behind every guess • Behaviour analytics over the signal time series: advertising interval, jitter, trend and periodicity • Radio tomography that maps excess path loss into an attenuation picture of the space FEATURES • Top-down radar with range rings, pinch to zoom, double tap to reset • Device list with search and type filters • Device card with signal history, ranked hypotheses and the evidence for each • Unwanted-tracker heuristics based on public advertising patterns • CSV export of a session • Full Russian and English, switchable inside the app PRIVACY All sensing and all math run on your iPhone. No account, no analytics, no tracking, no advertising identifier. One optional setting — off until you turn it on — may send a 24-bit manufacturer prefix to a public vendor database to name your router. Nothing else ever leaves the device. Use cool-radar only on networks and devices you are entitled to inspect. It does not decrypt Find My identities, does not enumerate neighbouring Wi-Fi networks, and provides no covert tracking of people. REQUIREMENTS iOS 18 or later. Bluetooth works on every iPhone. UWB distance needs U1/U2 hardware on both devices. Room geometry needs a LiDAR model. Anything unavailable on your hardware is shown as unavailable rather than simulated.