Ford Patent Proposes Cars That Move Themselves to Frustrate Thieves

A parked car is normally valuable precisely because it stays where its driver left it. A recently published Ford patent application reverses that expectation: After detecting what it classifies as “suspicious activity,” a vehicle could move itself to another location or change its orientation to make theft more difficult.

For anyone who has returned to find a vehicle missing its wheels, cargo or components, the appeal is immediate. The application, submitted to the United States Patent and Trademark Office, describes a car responding when someone repeatedly peers through its windows, when another vehicle circles the area or when a previously populated parking lot empties and leaves the car isolated. The unresolved tradeoff is equally clear: A theft-deterrence system based on unattended movement must also control where, when and under what conditions the vehicle moves.

Parking position becomes part of the security system

The proposal does more than send the vehicle elsewhere. It treats the surrounding terrain, walls and other parked vehicles as physical barriers that can deny access to valuable parts.

To protect its wheels, the car could park tightly between immovable objects or other vehicles. It could select an incline or soft or uneven ground where positioning and operating a jack would be more difficult. If equipped with air suspension, the vehicle could lower its body to reduce access to underbody equipment such as a catalytic converter.

Cargo protection would require a different orientation. The vehicle could back toward a wall until there was insufficient room to open its hatch or tailgate. Similar positioning could obstruct access to headlights or taillights. Ford’s concept would let the user choose whether the priority is protecting the wheels or securing the cargo area.

That choice illustrates the system’s practical limitation: One parking position may not protect every vulnerable area. Backing close to a wall can secure a tailgate but may leave the sides exposed. Restricting access to the wheels requires suitable objects, terrain or adjacent vehicles. A useful location therefore depends on both the vehicle’s hardware and the geometry available on an ordinary street or in a parking lot.

Detecting concern is different from authorizing motion

The proposed triggers also create a difficult perception problem. Looking through a window, circling a lot and approaching an isolated vehicle can be signs of suspicious behavior, but they can also occur for harmless reasons. A research project presented at the 20th Brazilian Symposium on Information Systems demonstrated that predefined suspicious behaviors near vehicles can be detected with computer vision, reporting 94.25% precision and 86.99% accuracy in its proof of concept. Those results show that behavior classification is technically feasible in a defined test setting, not that every ambiguous parking-lot encounter can be interpreted correctly.

Ford’s application would add a second decision after detection: whether the vehicle should remain still, recommend another location or relocate autonomously. That escalation matters because an alert can be dismissed by a driver, while physical movement changes the car’s position in a shared environment. The application leaves the resulting safety, control and regulatory questions unresolved.

A less complex implementation provides a more conservative path. The vehicle could recommend a safer location or parking orientation while requiring the driver to perform the maneuver. That preserves the idea’s main benefit using the surroundings to restrict access without giving an unattended car authority to leave its space. The cost is that the system cannot respond after the driver has walked away.

Ford has explored vehicle-movement concepts in other filings, including an external controller intended to help reposition a vehicle over short distances. That broader interest indicates that low-speed movement outside the conventional driver-in-seat arrangement is an area of intellectual-property work, but it does not establish a production plan for this anti-theft proposal.

This remains a patent application, not an announced Ford feature. Even receiving a patent would not mean the company will install the system in a production vehicle. Its most practical element may ultimately be the simplest one: advising a driver how to park against a wall or between barriers. Letting the car make that decision and then move after someone peers through its windows would buy a more active defense at the cost of a much harder safety and control problem.

By Robert McKinney — Editor-in-Chief for AMI’s automotive and mobility coverage, with a mechanical engineering background and a decade reporting on powertrain systems, EV innovation, and global vehicle manufacturing.