The concern is real. The most dramatic version of the claim is not always accurate.
SignalTrace is not a Flock product. It is sold by Leonardo US Cyber & Security Solutions under the ELSAG vehicle-recognition brand. The company previously called it EOC Plus.
It is not described as reading your messages, files or calls. Leonardo says it passively detects supported wireless signals that devices broadcast into the surrounding environment.
It can still create sensitive information. Leonardo says the system records electronic-device observations with date, time and location; groups signals that repeatedly travel together; stores the results for later queries; and can correlate them with license-plate observations when sensors and LPR cameras cover the same place.
That creates a path from tracking a plate to tracking a recurring combination of devices—and potentially following that combination when the plate is missing, changed or associated with another vehicle.
Leonardo: SignalTrace product page
Leonardo: SignalTrace product sheet (PDF)
What the names mean
Leonardo is the parent-company name used in current marketing.
Leonardo US Cyber & Security Solutions is the US subsidiary selling the system.
Selex ES is a legal/vendor name that can appear in contracts, bids and invoices.
ELSAG is the product brand used for Leonardo’s license-plate readers and related software.
EOC means Enterprise Operations Center, Leonardo’s platform for storing, searching and analyzing LPR and related information.
EOC Plus was the earlier name for the electronic-signal capability now marketed as SignalTrace.
A useful records search therefore needs all of those names. Searching only for “SignalTrace” could miss an older quote, pilot, invoice or contract labeled EOC Plus, ELSAG or Selex ES.
How SignalTrace is designed to work
Leonardo’s product materials describe strategically placed sensors observing supported wireless signals near a collection point.
The system can:
- detect broadcast signals from supported Wi-Fi, Bluetooth, RFID and vehicle devices;
- attach date, time and location to an observation;
- look for devices or device types that repeatedly appear together;
- treat that recurring group as an electronic signature;
- store and query the observations in the ELSAG Enterprise Operations Center;
- correlate a signature with an LPR observation when both systems observe the same place and time; and
- search for the signature again at other sensor locations.
Leonardo’s patent materials describe a vehicle carrying its own transmitting components while a person inside has a phone, smartwatch, access card and wireless headphones. The proposed system associates the observed signals with the vehicle’s visual identifiers, time and location.
US Patent 12,614,241: Systems and methods for electronic-signature tracking
What devices does Leonardo say it can observe?
The current product sheet lists categories including:
- mobile phones, tablets and laptops;
- smartwatches, fitness trackers and wireless headphones;
- Bluetooth devices and Wi-Fi sources;
- vehicle hotspots and infotainment systems;
- Bluetooth or RFID tire-pressure sensors;
- key cards, asset tags and other supported RFID tags;
- vehicle components and other supported security or safety sensors.
The product sheet also says “pet microchips.” Independent technical analysis questions whether that particular claim is realistic at roadside range and notes that Leonardo’s patent instead refers to pet collars. We therefore treat pet-microchip detection as a vendor marketing claim with unresolved technical support, not an established operating capability.
Plate readers and SignalTrace are different layers
| Traditional ALPR observation | SignalTrace observation | Potential combined record |
|---|---|---|
| Plate, vehicle image, time and camera location | Detected wireless signals or device characteristics, time and sensor location | A plate or vehicle associated with a recurring group of devices |
| Primarily follows a vehicle identifier | Can be deployed without an LPR and away from roads | May help follow a recurring signature when a plate changes, is obscured or is not captured |
| Does not by itself show which personal devices are nearby | Does not need to read device contents to recognize recurring signals or combinations | Can support movement, association and convoy analysis |
Leonardo says SignalTrace can operate with or without a co-located license-plate reader and can be used in non-traffic environments such as rail stations, shopping centers and event venues.
That matters. A roadside system can associate a device group with a vehicle. A separate sensor at a venue could later observe a similar signature without seeing the vehicle at all. Whether a particular deployment actually does this depends on sensor coverage, signal quality, software, configuration and investigative use.
What it does not appear to be
Based on the vendor and patent materials reviewed:
- it is not described as decrypting messages or reading stored files;
- it is not described as listening to call audio;
- it does not automatically obtain a person’s legal name from a phone;
- it is not the same thing as facial recognition;
- and it is not documented as a cellular IMSI catcher or “Stingray.”
We did not find support for saying SignalTrace captures every device or every signal that passes it. Detection depends on whether a device is transmitting a supported signal at the right moment, along with range, antenna placement, radio channel, traffic, interference and device privacy features.
Those limits do not make the collected data harmless. They mean the system’s conclusions can be incomplete or probabilistic rather than magical or perfectly precise.
Why “we do not read content” does not resolve the privacy issue
Content and metadata answer different questions.
SignalTrace may not read what a message says. A stored observation can still help answer:
- which device signature appeared here;
- when it appeared;
- where else it appeared;
- which other devices regularly appeared with it;
- which plate or vehicle appeared at the same time;
- and whether the group later appeared with a different vehicle.
Leonardo calls the signatures anonymous. That description deserves qualification. A signal may not contain a name, but once a recurring signature is associated with a plate, a home, an account, a known device or another identifying record, it can become linkable to a person.
NIST guidance treats information as personally identifiable when it can distinguish or trace someone, including when seemingly non-identifying information is linkable to another source. A plate plus repeated device, time and location observations presents exactly that kind of combination risk.
NIST: Guide to Protecting the Confidentiality of Personally Identifiable Information
The main civil-liberties risks
Movement tracking beyond one plate
A recurring electronic signature offers another way to recognize a vehicle or group even when a plate is changed, missing or misread. Repeated sensor coverage could make routine travel searchable through the devices that tend to accompany it.
Association mapping
Leonardo markets detection of signatures that frequently travel together and convoy patterns. That same capability can expose recurring associations among family members, coworkers, friends, passengers, organizers or people attending the same event.
Pedestrian and venue surveillance
Because the sensors can operate without an LPR and in shopping centers, stations or event venues, the surveillance model is not limited to cars. Devices carried by pedestrians and bystanders may enter the collection area too.
Innocent-device and innocent-vehicle association
A friend, rideshare driver, borrowed vehicle, crowded intersection or nearby pedestrian can place unrelated devices near a target observation. If a system treats a recurring correlation as identity, investigators may infer a relationship that has an innocent explanation.
Sensitive-place inference
Repeated observations can expose visits to medical offices, houses of worship, legal services, union activity, political events, protests or private homes without recording what happened inside. The location and association can be revealing on their own.
Function expansion without visible replacement
Signal sensors, LPR cameras and EOC software are separate components. New licenses, integrations, collection points or data sources can expand a system without every change being obvious from the street. Residents need product inventories, entitlements, configuration histories and audits—not only a photograph of a pole.
SignalTrace also has meaningful technical limits
Modern Apple and Android devices use forms of Wi-Fi address randomization intended to reduce tracking. Bluetooth devices can also rotate identifiers. Devices transmit intermittently, use different channels and may pass through a sensor’s useful range in less than a second.
Apple: private Wi-Fi addresses
Android Open Source Project: MAC randomization behavior
Independent wireless-security analysis explains that a roadside antenna cannot perfectly assign every detected signal to one passing vehicle. Nearby cars and pedestrians may blend into the observation, while other devices may not transmit during the brief collection window. Repeated observations and confidence scoring can improve a correlation, but they do not turn inference into certainty.
Ryan O’Horo: technical analysis of SignalTrace
The accurate criticism is therefore not “it flawlessly identifies every phone.” It is:
The system is designed to collect and retain wireless-device observations about everyone within range, then use repeated correlations to infer movement and associations. Some people will be missed; some signals may be associated incorrectly; and both outcomes create risk.
The legal line is not settled specifically for SignalTrace
In June 2026, the US Supreme Court held in Chatrie v. United States that obtaining Google’s stored location-history data through the geofence process at issue was a Fourth Amendment search. SignalTrace is technically and factually different because it involves direct sensor collection of broadcast signals rather than demanding stored Google records.
We have not found a reported court decision deciding whether a particular SignalTrace collection or search requires a warrant. Officials should not treat “broadcast in public” as the end of the legal analysis when the system is designed to aggregate location and association data over time.
Supreme Court opinion: Chatrie v. United States
This page provides general public information, not legal advice.
Is SignalTrace operating in Metro Detroit?
We have not verified a SignalTrace or EOC Plus deployment in Metro Detroit as of September 29, 2026.
That is a statement about the records found so far—not proof that no agency has tested, purchased or accessed the technology.
Our review of indexed official material for Detroit and selected Wayne, Oakland and Macomb County communities did not locate a contract, invoice, policy, agenda item or equipment inventory naming SignalTrace or EOC Plus.
Detroit provides one important procurement lead. The city’s May 21, 2024 contract report lists Selex ES, Inc. among eight bidders for Detroit’s freeway-ALPR project. Selex received 17 evaluation points. The award ultimately went to Flock Group, not Selex. That document proves Leonardo’s legal subsidiary participated in the procurement; it does not prove Detroit bought Leonardo hardware, ELSAG software or SignalTrace.
Detroit City Council contract report, May 21, 2024 (PDF)
DPD’s March 2026 surveillance response names Flock, Motorola and Genetec LPR systems, not Leonardo or SignalTrace. Neither document answers whether Detroit evaluated EOC Plus inside the losing Selex proposal, used a demonstration or later obtained a separate product.
DPD data-sharing and surveillance response, March 10, 2026 (PDF)
The correct local status is therefore:
SignalTrace capability: documented and actively marketed.
Metro Detroit deployment: not currently verified.
Detroit Selex/Leonardo contact through bidding: verified, but not an award or deployment.
Do not identify SignalTrace from appearance alone
Leonardo sells ordinary ELSAG plate readers that do not, by that fact alone, establish SignalTrace collection. SignalTrace also uses sensor hardware that may not resemble a familiar camera.
A crowdsourced map tag or street photograph is a useful lead. It is not enough to prove the product, enabled capability, data flow or operator. Confirmation should come from contracts, equipment inventories, serial/model records, network diagrams, policies, configuration exports, invoices or a direct written agency answer.
The records that can establish local use
Ask for existing records containing any of these names:
- Leonardo US Cyber & Security Solutions;
- Leonardo;
- Selex ES;
- ELSAG;
- SignalTrace;
- EOC Plus;
- Enterprise Operations Center or ELSAG EOC;
- electronic-signature detection;
- Wi-Fi, Bluetooth, RFID or radio-frequency collection associated with LPR;
- Field Control Unit or FCU associated with Leonardo/ELSAG;
- and any integration connecting those systems with Flock, Motorola/Vigilant, Genetec, a real-time crime center or another evidence platform.
Request:
- contracts, order forms, quotes, invoices, grants, purchase orders and bid responses;
- product and license inventories, including trials, demonstrations and no-cost pilots;
- equipment models, installation sites and responsible operating agencies;
- data dictionaries showing every collected field;
- retention, deletion, export, preservation and backup settings;
- user roles, search permissions, sharing partners and audit logs;
- policies, training, legal reviews and warrant requirements;
- integration diagrams, data-flow documents and API/interface descriptions;
- validation, accuracy, false-association and confidence-scoring records; and
- records showing when a feature was enabled, disabled or renamed.
Ask for an estimate before fees exceed a reasonable amount, and coordinate locally so residents do not send burdensome duplicate requests.
Use our Michigan public-records guide →
Questions officials should answer before any purchase or pilot
- Why should the government collect device observations from everyone within sensor range when most people are not suspected of a crime?
- What exact signals and identifiers are recorded?
- Can observations be linked to plates, addresses, names, accounts or other databases?
- Can the system alert on a recurring electronic signature?
- Can it follow devices away from vehicles or inside venues?
- How are bystanders and multiple nearby vehicles separated?
- What confidence threshold is required before a lead is acted upon?
- Is a warrant required for historical searches, prospective alerts or association analysis?
- How long is raw and derived data retained, and what happens to exports?
- Which governing body must approve new sensors, integrations or signal types?
- How will the public learn about a change in capability?
- What is the termination, return and verified-deletion process?
Our position
Not reading message content does not make automatic location and association tracking benign.
SignalTrace is designed to turn ordinary wireless emissions into a searchable investigative dataset. It can collect first, identify later and infer relationships from the devices that travel together. Technical limits may make it less reliable; they do not make mass collection more acceptable.
Metro Detroit governments should not purchase, pilot or activate SignalTrace or a materially similar device-signature system. Any existing deployment should be disclosed, suspended and brought to a public decision for termination, removal and verified data closeout.
The best time to stop this infrastructure is before it becomes another quietly renewed line item.
Research the system in your municipality →
Understand how access, sharing and retention expand a network →