Opinion: Tesla’s Robotaxi Puts Company In Ludicrous Mode

Share on facebook
Share on twitter
Share on reddit

Tesla’s racy “Ludicrous mode” has become a signature of the Silicon Valley disruptor and its equally fast-moving EV’s. But Elon Musk’s otherwise laudable race to be first in all things tech—electric cars, a fast-charging network, private space transportation—has produced a truly ludicrous claim: That by next year, with the flip of a software switch, Tesla will unleash 1 million “robotaxis” on roads. These electric cars would become 24/7 moneymakers, ferrying paying customers with no human behind the wheel. As for who’ll be responsible for shampooing Teslas’ back seats after they’re defiled by puking frat boys, Musk wasn’t saying.

This million-strong taxi army was news to every self-driving competitor, from Google’s Waymo to General Motors’ Cruise unit. Those companies are proceeding on a vastly more-cautious timeline, convinced that full self-driving isn’t ready for prime time, let alone 3 a.m. on a deserted back road. Even Musk sensed that his “Autonomy Day” announcement was a doozy, telling investors that “Sometimes I am not on time, but I get it done.”

Tesla and Musk have earned some wiggle room here. This is the company that created the market for EV’s. While other automakers paid lip service to green cars and climate change, Tesla did something about it. Tesla still struggles to rub two nickels together in terms of profit. But it’s selling more EV’s by far than any company on the planet, including roughly 230,000 in 2018, and perhaps more than 400,000 this year.

But while Tesla has delivered on some promises, it has broken others, including a bait-and-switch on the $35,000 version of its Model 3 that would have been a PR catastrophe for any other car company. This latest promise stretches credulity to the breaking point, raising questions from technical feasibility and legal liability to political, social and regulatory censure that Musk hasn’t begun to address.

Here’s my promise: Tesla will put no more than a handful of ride-sharing, driverless cars on the street next year, or by 2021, or 2022. I’m talking all types of streets as pledged, not just the highways currently plied (very well, admittedly) by its deceptively named “Autopilot.” Tesla’s vaunted new self-driving microchip sounds awesome, but autonomous experts I’ve spoken with say there’s no way it’s ready to safely manage the complex ballet of driving, in any and all conditions. Consider Cadillac Super Cruise, a system that in some ways has already trumped Autopilot. Super Cruise is largely hands-free, but it’s far from human-free, and that’s the point. It integrates detailed, Lidar-based mapping. It’s geofenced to operate on divided highways only. It’s even got an infrared, driver-monitoring camera to keep pilots engaged, and alert them to retake control when necessary. Most autonomous experts see those three features as critical to the safe implementation of self- and semi-autonomous driving. Tesla and Musk intend to proceed with precisely none of them.

Instead, Teslas will be limited to the same basic radar and cameras (the latter front-facing only) that you’ll find on any Hyundai or Subaru. Yes, Tesla has the most real-world autonomous testing under its belt, and some of the best deep-learning knowledge in the game. Yet we’re to believe that its current models, upgraded with a fancy computer chip, will chauffeur passengers in fail-safe fashion, from a nighttime snowstorm in Montana to the tourist hordes of Times Square? Please. Musk talks about how humans drive using exclusively vision, but fails to mention how that vision is hooked to the peerless, bioengineered computer that is the human brain. In my extensive experience with self-driving and semi-autonomous cars—including racing in and against Audi’s autonomous RS7 on a Spanish track—I’ve seen cameras flummoxed by something as simple as shadows on a bright day. Sensors can fail, which demands more hardware redundancy than Tesla cares to deliver at this point.

Elon Musk claims that with this computer and a basic suite of sensors, Tesla vehicles will be capable of autonomous driving.

 

There’s another practical bar to converting Tesla’s expensive, tech-laden EV’s into virtual Ubers, minus the poorly compensated drivers: They were never designed for this application. Starry-eyed fleet operators have tried to enlist Teslas as taxis, with disastrous results that have sparked recriminations, bankruptcies and legal wrangling, from Canada to Sweden to China. Operators have found that Teslas simply can’t handle the brutal stresses of daily taxi operation. The broken-down Teslas then sit idle for days or weeks because of a shortage of replacement parts and repair techs. There’s a reason so many Uber drivers choose Toyota Camrys and other affordable, bulletproof, easy-to-service models. Long EV charging times are another issue for owners who buy into Musk’s fuzzy math, the claim that their cars will earn them up to $30,000 a year in gross profits, even as they sleep.

Let me be clear: I’m bullish on the potential for both self-driving cars and Advanced Driver Assistance Systems (ADAS) that amplify, rather than replace, human skills. These technologies can and will reduce injuries and fatalities, unclog traffic, save precious time and be a godsend for elderly, disabled or inexperienced drivers—someday. But Musk tends to get ahead of himself, distracted by his latest tech vision and competitive self-interest. We’ve seen three Tesla fatalities linked to Autopilot’s literal blind spots, and an pedestrian killed by an autonomous Uber prototype. Those highly publicized accidents all occurred with a human still behind the wheel . Many consumers are already skeptical, or unnerved by the idea of driverless cars. Imagine their reaction when the first paying customer is injured or killed in a driverless Tesla that wasn’t as foolproof as advertised— perhaps those aforementioned, tipsy college students, stuck in the backseat and powerless to avert the crash.

It’s possible that Musk himself doesn’t believe the dubious timeline he floated. Perhaps “Autonomy Day” was just another well-timed shot of PR adrenaline to boost Tesla’s flatlining stock. But judging from Tesla investors’ blase reaction—the company’s stock price barely budged—I’m not the only one who’s not onboard for the robotaxi fantasy.


About the Author

Close Menu