Research-driven solutions that protect consumer privacy, safety, and autonomy.

Vulnerability disclosure

We examine widely deployed consumer devices for security weaknesses and report what we find to the manufacturer, so it can be fixed before it is exploited.

The work is unglamorous and mostly invisible. It matters because the people using these devices have no practical way to evaluate them, and no recourse when something is wrong.

Passive radio detection and signal correlation

Nearly every powered device now broadcasts something — a beacon, a probe, a periodic signal. Those transmissions are public, unencrypted, and continuous. B-Hound is a passive receiver and correlation engine that reads them and answers a deceptively simple question: which of these signals has been travelling with this person, or lingering in this place, for longer than it should?

The method is general because the question is. Distinguishing a device that follows from one that is merely present is the same computation whether the subject is a tracker in a bag, a sensor on a job site, or equipment that has left a building. What changes is which answer matters.

Applications under evaluation:

• Personal safety — detecting hidden trackers and covert recording devices
• Family and caregiving — consented awareness for households and memory care
• Asset and equipment tracking — locating tools, inventory, and hardware without a fixed installation
• Worksite safety — proximity awareness and evacuation accountability
• Counter-surveillance auditing — surveying a space for devices that should not be present
• Aggregate measurement — occupancy, flow, and dwell analysis with no individual record retained

The same signals that make surveillance possible make detection possible. The difference is what happens next: surveillance stores what it sees; detection does not. That distinction is architectural, not a matter of policy.

Learn about B-Hound →

Quantitative research and consulting

Statistical consulting, dissertation and research support, curriculum design, and corporate analytics training for institutions, researchers, and teams.

This is the methodological backbone of the group. Detection is a classification problem before it is a hardware problem, and the standards that govern peer-reviewed research — study design, error control, reproducibility — are the same ones that decide whether a safety product can be trusted.

Visit Green Line Analytics →