A drawing accompanying Ford's patent application for an autonomous police vehicle, last updated on U.S. Patent and Trademark Office's website on Jan. 18. (Image: Courtesy of the U.S. Patent and Trademark Office)

The police cruiser sets up a speed trap, training its radar at oncoming traffic. It soon hits on a car going way too fast. As it pulls away from the curb, the computer aboard the cop car communicates with the target car, pulling up the registered car’s driver’s license—and determining whether it is being driven manually by a human hand or by autonomous steering.

Because the police car is automatically engaging this pursuit, as Ford envisions in a recent patent.

Ford Global Technologies, based in Michigan, filed the patent for the “autonomous police vehicle,” and it was most recently updated on Jan. 18.

Machine learning will even pick the “good hiding spot” for where to set up the speed trap in the first place, according to the patent filing.

“Routine police tasks, such as issuing tickets for speeding or failure to stop at a stop sign, can be automated so that human police officers can perform tasks that cannot be automated,” the engineers write in the patent, originally filed in 2016 but dated Jan. 18, 2018. “Accordingly, the present disclosure describes autonomous police vehicles that can, on behalf of human police officers, perform automated tasks such as enforcing traffic laws and issuing tickets/citations to drivers that violate the traffic laws.”

Many technology onlookers believe that autonomous vehicles will be dominating the roads of America in the coming decade. Of course, there have been hurdles in their development, which is undertaken by most of the major automakers. 

The patent, which was originally reported by the site Motor1, describes a mix of sensors and artificial intelligence that will make the autonomous vehicle able to make some judgment calls currently made by human eye and hand.

For instance, when a car is behaving suspiciously, the autonomous car may “assume” a drunk driver is behind the wheel.

“When the movement of (the target car) appears suspicious (e.g. sudden stop, meandering movement, abnormal lane changes or the like), the autonomous police vehicle may pull it over under the assumption that (the target vehicle) is in manual driving mode and that the driver is under (the) influence,” it describes.

A police officer, either aboard the autonomous cop car or in the nearby area, would then have to follow up to determine whether the driver is indeed under the influence, they explain.

Another scenario at a roadside stop: whether to ticket, or simply warn the driver. The machine would be making the call.

“The autonomous police vehicle may wirelessly transmit a second message to (the target vehicle), with the second message indicating the disposition as a result of the determination,” the patent says. “For example, the disposition may include a ticket with a fine or a warning without a fine, and may also include a message indicating that (the target vehicle) is free to leave the scene.

“In determining whether to issue a ticket or warning, (the) autonomous police vehicle may search a local record or query a central computing system to look up any record of violation of traffic laws by (the target vehicle) or any driver associated with (it),” it adds.

Automated technologies at use in police work have been controversial. For instance, red-light cameras have proven extremely unpopular in places like New Jersey, where they have been discontinued. However, the use of automated license plate readers (ALPRs) has been widely expanded to include most states—though their automated scanning of license plate and driving histories only supplements a patrol officer’s presence behind the wheel.