Skip to Content
Artificial intelligence

How a College Kid Made His Honda Civic Self-Driving for $700

Who needs a Tesla when you can build your own automated copilot using free hardware designs and software available online?
February 21, 2017

Brevan Jorgenson’s grandma kept her cool when he took her for a nighttime spin in the Honda Civic he’s modified to drive itself on the highway. A homemade device in place of the rear-view mirror can control the brakes, accelerator, and steering, and it uses a camera to identify road markings and other cars.

“She wasn’t really flabbergasted—I think because she’s seen so much from technology by now,” says Jorgenson, a senior at the University of Nebraska, Omaha. Others are more wary of the system, which he built using plans and software downloaded from the Internet, plus about $700 in parts. Jorgenson says the fact that he closely supervises his homebrew autopilot hasn’t convinced his girlfriend to trust the gadget’s driving. “She’s worried it’s going to crash the car,” he says.

Many tech and auto companies have begun testing modified cars on the road in recent years. Jorgenson’s vehicle is in the vanguard of a more ragged, grassroots test fleet taking shape as tinkerers around the world strive to upgrade their own vehicles with computing gear that can share driving duties.

Motivation comes from the fun and challenge of getting the technology working—and the prospect of making driving easier. Kiki Jewell, who set out to make her Chevy Bolt self-driving as a learning exercise, says her spouse has been strongly supportive, partly out of self-interest. “My husband’s happy I’m interested to ease his commute,” says Jewell, who lives in the Bay Area.

Jewell and Jorgenson’s projects were enabled by a fit of pique last October by the founder of Comma.ai, a San Francisco startup that was developing a $999 device that could upgrade certain vehicles to steer themselves on the highway and follow stop-and-go traffic. Founder George Hotz abruptly cancelled plans to launch the product after receiving a letter asking questions about its functionality from the National Highway Traffic Safety Administration. In November, he released the company’s hardware designs and software for free, saying he wanted to empower researchers and hobbyists. (Hotz didn’t respond to requests to talk about his strategy.)

College senior Brevan Jorgenson’s car can steer itself on the highway thanks to this device he built and installed using $700 in parts.

Jorgenson set about ordering the parts needed to build up Comma’s device, the Neo, the same day Hotz dumped the plans online. He had been following Comma’s fortunes, and he happened to own a 2016 Honda Civic, one of the two models supported by the company’s software (the other is the 2016 Acura ILX).

A Neo is built from a OnePlus 3 smartphone equipped with Comma’s now-free Openpilot software, a circuit board that connects the device to the car’s electronics, and a 3-D-printed case. Jorgenson got the case printed by an online service and soldered the board together himself.

He first put his life in the device’s hands in late January after an evening college class. “It was dark on the interstate, and I tested it by myself because I figured if anything went wrong I didn’t want anybody else in the car,” says Jorgenson. “It worked phenomenally.” Subsequent tests revealed that the Neo would inexplicably pull to the right sometimes, but a software update released by Comma quickly fixed that. Now fully working, the system is similar in capabilities to the initial version of Tesla’s AutoPilot (see “10 Breakthrough Technologies: Tesla AutoPilot”).

Comma’s plans and software aren’t the only resources out there for wanna-be self-driving-car builders. Neodriven, a startup based in Los Angeles, recently started selling a pre-built Neo device that works with Comma’s Openpilot; it costs $1,495. Online-education platform Udacity has released code used in its autonomous-car research program, and students in one of its courses are actively improving and expanding it (“The Creator of Google’s Self-Driving Car Now Competes with It”).

Bryant Walker Smith, a law professor at the University of South Carolina, says that federal and state laws probably don’t pose much of a barrier to those with a desire to upgrade their vehicle to share driving duties. NHTSA has authority over companies selling vehicles and systems used to modify them, but consumers have significant flexibility in making changes to their own vehicle, says Smith, who advises the U.S. Department of Transportation on law and automation.

Anyone using a home-built Neo will still have to comply with state rules requiring responsible driving, though. (Comma’s Openpilot software tries to help with that: it complains if the driver doesn’t touch the wheel every five minutes, and it asks for human intervention if it’s having trouble interpreting the road ahead.) And in the event of a crash, using a home-built driving aid might raise eyebrows. “Just because you can legally operate it doesn’t mean you are not civilly liable,” says Smith.

Ariel Núñez, a software developer in Barranquilla, Colombia, hopes the work of hobbyists like himself will show how existing cars could be made significantly safer—an alternative vision to that chased by giant companies focused on ending the need for human drivers. He’s using code from Comma and Udacity to try to get his Ford Fusion to automatically slow down when it sees traffic signs, speed bumps, or potholes (he hasn’t tested it on the road but has got the accelerator and steering control working, and had a near miss with a tree). “I am less interested in full autonomy and more in preventing rear-endings,” he says. “A lot of existing cars can be retrofitted.”

Deep Dive

Artificial intelligence

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

Providing the right products at the right time with machine learning

Amid shifting customer needs, CPG enterprises look to machine learning to bolster their data strategy, says global head of MLOps and platforms at Kraft Heinz Company, Jorge Balestra.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.