scorecardresearch Skip to main content

We took a ‘self-driving’ Tesla for a spin in Boston. Here’s how it did.

We took a ‘self-driving’ Tesla for a spin in Boston. Here’s how it did.
“It’s unbelievable that people will say that ‘this is safer than a human driver,’” Boston driver and Tesla Model Y owner Taylor Ogan said. “That’s not true.” (Video by Pranshu Verma, Edited by Caitlin Healy/Globe Staff, Photo by David L Ryan/Globe Staff)

I will confess: Until this week, I’d never sat in a Tesla. So imagine my excitement on Wednesday, when I got invited to ride in one. For all the wrong reasons.

It all started with a tweet from a Boston driver named Taylor Ogan. He drove his Black Tesla Model Y on self-driving mode around South Boston, showing in a video how it fails to navigate city streets properly. This was interesting, given how impassioned Tesla drivers can get in extolling the benefits of the company’s various self-driving features. And given the technology’s ongoing safety and regulatory issues.

I reached out to Ogan on Twitter. Twenty-four minutes later, he responded. “Yes,” he wrote. “We can do an interview in the car so you can experience it, too.”


When we met in downtown Boston and I got into his jet-black Tesla SUV, equipped with white leather seats, I learned a few things. “Full self-driving” is a feature in beta testing that Tesla drivers of certain new models can pay upwards of $12,000 to unlock. It comes in three modes: chill, average, and assertive. The feature rolled out in mid-January and is different from Tesla’s autopilot feature, which requires more effort and attention from the driver. (A little over 53,000 cars in America are equipped with the full self-driving feature, news reports show.)

At first, we decided to give the car’s “assertive” mode a try. That allows the vehicle to perform rolling stops, a feature that is currently being recalled, according to the US Department of Transportation.

Pretty soon, things started going awry. At one intersection in South Boston, where multiple lanes converge, we inched our way forward and then stopped in the middle of the street, drawing honks from nearby cars. Shortly after that, we veered unexplainably into a snow bank, forcing Ogan to take control of the wheel to stop us from denting his car.


Soon after, we took the car off aggressive mode, and gave the “average” setting a try. Still, problems remained. At one point, a van was parked in the middle of the street, and leaving the Tesla to its own devices, it just sat there, thinking what to do. Ultimately, Ogan had to steer around it.

“It’s unbelievable that people will say that ‘this is safer than a human driver,’” Ogan said. “That’s not true.”

So, does the company need better technology for its self-driving mode? And where do we go from here?

Currently, Tesla’s self-driving feature relies on cameras to analyze the world around it. Those video feeds get sent to the car’s onboard neural network, which analyzes it for roads, cars, obstacles, and people. But, experts have said, that technology can be inaccurate in detecting certain objects. That is why many self-driving cars, such as Waymo’s, use LIDAR technology, which emits laser beams to create 3D maps and supplement the gaps in computer vision.

Some electric car enthusiasts, like Ogan, believe Tesla would benefit from putting LIDAR into its cars. Elon Musk has previously referred to the technology as a “crutch.”

Either way, Ogan, a hedge-fund executive who owns Tesla stock, believes with the self-driving tech in its current state, it is unsafe for the feature to be available to the general public. He said Tesla should follow the lead of other companies, such as Boston-based Motional, and rigorously test the feature with company experts driving, in a cordoned-off section of a city, with permits and safety protocols.


“You know why?” he said. “Because human lives are at risk here.”