Browse
Log in / Register
Share
July 13, 2026 — Police surveillance is gaining a new layer. Leonardo's SignalTrace, a security system designed to work alongside automatic license plate readers, can match wireless device signals to specific vehicles, giving investigators a way to develop leads before they know a driver's name. The system is field-tested in Maryland and appears on a New York state contract price list. Privacy scholars are beginning to ask how it squares with Fourth Amendment protections.
Picture a routine morning commute. Two co-workers pass the same roadside sensor suite each day, and a plate reader logs their car, pulling registration data. A second sensor inside the same housing picks up Bluetooth and Wi-Fi probes from devices in the vehicle — a phone in the driver's pocket, a smartwatch on the passenger's wrist. After a few trips, the system detects that those two electronic signatures reliably travel together. That pattern is the foundation of SignalTrace: consumer electronics form a recognizable signature swarm around a vehicle.
Leonardo, an Italian defense contractor with a major U.S. footprint, pitches the product as an investigation aid rather than a tool for identification. Its explanatory materials emphasize that the system "does not identify people" and that it merely collects the electronic signatures devices already broadcast — no content, no names, no account records. The company says its output must still be corroborated through ordinary detective work.
That framing is accurate in the narrowest sense, but misleading in practical terms. A Bluetooth identifier does not carry a legal name, and it does not need to. If the same signature keeps appearing in a car registered to one individual, outside that person's home, or next to another device already known to police, the carrier can be identified by inference. At that point, the "nameless" signal becomes a personal identifier in every practical way.
The privacy concern centers on what happens before a name is attached. Movement patterns alone can be surprisingly identifying. A study in Scientific Reports examined 15 months of mobility records covering 1.5 million people and found that four time-location points were enough to uniquely identify 95% of them. SignalTrace was not tested in that research, but the math applies directly: sequences of time-stamped locations are inherently distinctive.
Courts are starting to recognize that reality. In Carpenter v. United States, the Supreme Court ruled that the government's acquisition of cell-phone location records violates a reasonable expectation of privacy and therefore constitutes a search. This June, in Chatrie v. United States, the Court extended that logic. Police investigating a bank robbery had used a geofence order to obtain anonymized location data for phones near the bank, narrowed the pool by movement patterns, and only later attached names to specific users. The Court held that the initial location-data acquisition — name or no name — was a Fourth Amendment search.
Chatrie does not settle SignalTrace's legality, but it frames the open question. The geofence records in Chatrie came from cell carriers; SignalTrace captures signals directly from roadside hardware. The Court's underlying concern, however, was about the state quietly amassing a detailed record of where people go. SignalTrace adds a wrinkle: none of its raw data reveals a subscriber's identity, yet it is designed to connect that data to plate records and produce identifying leads. Whether the courts treat that as a search remains unresolved.
Association is a second layer of risk. When a set of device signatures repeatedly clusters, SignalTrace can treat the cluster as a recurring electronic pattern. It could be a family car, a carpool, or a group of activists driving to the same events. The system does not understand the relationship; it records co-occurrence. If one member of that cluster later becomes a target, every other device in the cluster becomes a lead. Researchers have shown how powerful such inference can be: a study in the Proceedings of the National Academy of Sciences tracked 94 participants with phones that logged Bluetooth proximity and found that off-hours contact alone classified 95% of mutually reported friendships. SignalTrace has not been independently evaluated to that standard, but the field research strongly suggests that repeated closeness reveals social ties.
Errors are also possible. Phones get borrowed, lost, or left in cars driven by someone else. Roadside sensors capture pedestrians walking past the antenna. A device-vehicle match does not establish who carried the device on a particular day. Leonardo markets SignalTrace as a lead-generation tool, and a lead has consequences: it directs officers toward certain people, records, and surveillance requests while leaving others untouched.
The system's architecture is worth noting. Leonardo's product page says SignalTrace stores electronic fingerprints for future queries and can recognize a vehicle without reading its license plate. A related patent describes targets that may be people or vehicles, and mentions searchable signatures that can be correlated with visual identifiers and used to track a target across locations. In other words, the design anticipates ongoing tracking, not merely one-time tips.
That changes the shape of an investigation. Traditional policing starts with a known person or incident, then gathers location evidence. SignalTrace inverts the sequence. Investigators can begin with a nameless electronic recurrence, form a hypothesis about who is behind it, then consult vehicle registrations, warrants, and subscriber records to attach identity. The original pattern looks like a passive observation, but it functions as an accusation dressed as data.
Legal definitions of personal information already cover this territory. The National Institute of Standards and Technology defines personally identifiable information as data that can distinguish or trace an individual's identity, alone or in combination with other information. A device signature paired with time and place can do exactly that. The complication is that most police-technology policies still assume identifying data arrives with a name attached.
License plate readers, of course, are nothing new. They generate tens of millions of vehicle images every day through police cruisers, toll gantries, and parking enforcement. SignalTrace's significance comes from convergence. Pairing plate data with wireless signatures creates a searchable archive of movement and association that did not exist a decade ago — an archive that current ALPR regulations may not cover.
State and local agencies now weighing SignalTrace face a policy gap. Several jurisdictions restrict how long ALPR data may be kept and who may query it, but those rules were written for plate metadata, not device proximity patterns. Whether they extend to a companion database of electronic signatures remains untested.
For now, SignalTrace occupies a gray zone between demonstrated technology and settled law. Installations are reported in Oxon Hill, Maryland, and predecessor technology appears in New York procurement documents. As deployment grows, the distance between what the system can observe and what courts will allow will narrow — and the resolution will determine how much of everyday proximity becomes part of a permanent police record.
