What your premises tell the street: Wi-Fi and Bluetooth
Field: premises security, surface audit, open source Android app
Period: September 2026
Client: internal project, Lab 77 (R&D)
A Wi-Fi access point, a printer, a camera, a connected thermostat: each of these devices constantly announces its presence to anyone who walks past the building with a phone. Nobody decided this; it is simply how they work. I wrote a small open source Android app that records, without touching anything, what is broadcast in this way. It serves as a first surface audit: seeing your premises as the street sees them.
Context
For the security of a building, one question arises: what, from the outside, gives away what is inside? For radio waves, the answer exists: Wi-Fi beacons and Bluetooth advertisements, sent in the clear so that other devices can connect. Recording them requires no hacking and no special equipment, just a phone and software that is honest about what it does.
Problem
- Forgotten devices: an old access point left in place, a printer exposing its own network, a connected object installed without telling anyone.
- Names that say too much: a network or device name can reveal the company, the department, the brand of a piece of security equipment.
- Ill-suited tools: Wi-Fi diagnostic apps are aimed at troubleshooting a network, not at inventorying what a place broadcasts. Other tools go much further than necessary, and than permitted.
Solution
1. Listen, never query
The app is strictly passive. It receives what devices broadcast publicly, timestamps it and places it on the map. It intercepts no traffic, connects to no device, sends nothing. The network is only used to download map tiles, and these can be downloaded in advance to work entirely offline. No telemetry.
2. Classify by manufacturer, without claiming certainty
Every hardware address starts with a prefix assigned to the manufacturer. The app relies on the public table of these prefixes, taken from the Wireshark project, to classify each device: access point or router, network infrastructure, camera, professional connected object, consumer device. It is a deduction by manufacturer, never a certainty, and the app says so.
3. Show uncertainty instead of hiding it
Signal strength gives an idea of distance, never of direction: the display does not pretend otherwise. Only fixed equipment appears on the map.
4. Export for analysis
The survey exports in open formats (CSV and JSON) and as a text debrief.
Result
Version 0.3, released in September 2026, is the reliable base I set after several rapid iterations. It is waiting for field feedback before the next steps: the three languages, then publication in an open source app catalogue.
What I take away
- Stop shipping in the dark. I released thirteen versions in a single day. Each one fixed a detail; none made the tool reliable. I stopped everything to write a base: no position recorded without a reliable fix on the observer, an accuracy budget of 5 metres indoors and 10 metres outdoors, and one batch of changes per version.
- The phone already protects privacy. Most phones regularly change their Bluetooth address: the manufacturer cannot be deduced, and that is as it should be. What the audit sees is mostly fixed equipment, in other words the equipment you are responsible for.
- The system sets its limits. Android limits the number of Wi-Fi scans an app can start, to four every two minutes. An honest tool shows when it is throttled, instead of suggesting the list is complete.
The legal framework
Receiving what a device broadcasts publicly is not intercepting a communication. But the tool is meant for your own premises, or those of a client who has asked you in writing. Identifiers recorded at third parties are not published, and the law varies from country to country: check before recording anywhere other than your own premises.
Going further
Do you want to know what your premises reveal, and fix what needs fixing: forgotten access points, overly explicit names, connected objects on the wrong network? I carry out this first survey and help you put things in order. See your tools and your data, or book a 20-minute discovery call.
Sources
- Wi-Fi scanning overview, Android documentation, accessed 15 November 2026.
- MAC randomization behavior, Android documentation, accessed 15 November 2026.
- OUI Lookup Tool, Wireshark project, accessed 15 November 2026.
- Photo: Maripo GODA / Wikimedia Commons (commons.wikimedia.org), licence CC BY-SA 3.0.
Tags: premises security, Wi-Fi audit, Bluetooth, passive survey, Android app, open source software, connected objects