Are Waymo’s soon-to-launch self-driving taxis still undercooked?
Haters reveal Waymo’s woes, Jaguar Land Rover sees to pedestrians’ safety concerns and Dyson’s radical ramp-up: read our weekly analysis of the most important news in automated driving!
“I’ve done fully driverless in Phoenix as well a few times, and it’s pretty normal. It just works.” That’s what Ellice Perez, Head of Operations at Waymo recently told The Verge when it was granted exclusive access to Alphabet’s 70,000-square-foot self-driving depot in Suburban Arizona. But another exclusive published last week on The Information would beg to differ, claiming that the self-driving Chrysler Pacifica Hybrid minivans that Waymo plans to make available to the paying general public this year, are struggling. Big time.
Amongst several reported issues is something The Information refers to as ‘The Zoolander problem’: the vans really struggle to turn left. More specifically, unprotected left turns – when there is no automatic right of way or no left-turn lane. Other challenges include merging into heavy traffic on highways or into turn lanes, understanding basic road features like metered red and green lights at intersections and last but not least, reacting appropriately to drivers or pedestrians that fail to observe traffic laws down to the letter (e.g. speeding, not coming to a complete stop, illegal turns).
As a result, the vehicles apparently drive in an overly cautious manner (sudden stops, driving in the center of the road…) which, in turn, is infuriating local residents. “I hate them,” was uttered by more than a dozen when interviewed; “I’m sick of them in my neighborhood,” one said, while another admitted to illegally driving around them.
This, to me, begs the question: is everything really going as swimmingly as Waymo would have you believe? What’s most concerning about this report is not the driver frustration at overly cautious behavior (although that also does nothing to help the acceptance case), but the technical challenges these vehicles still face. As it so often boasts, Waymo has clocked 8 million miles of real-world experience. Surely the public are starting to ask: how many miles do you need to be able to learn a left turn?
The secret’s in their eyes
As I said, this sort of publicity doesn’t help regain peoples’ trust in automated driving. It’s up to the industry to make the public feel safe again – and Jaguar Land Rover UK thinks it may know how.
The British firm has fitted human-like eyes to the fronts of autonomous cars to give pedestrians a sense of security that the car has recognized them and in doing so gives them a signal to cross. The pods with eyes are being put to the test with volunteer pedestrians in the safety of a fabricated street scene in a facility in Coventry, UK.
As Pete Bennett, Future Mobility Research Manager at Jaguar Land Rover, says: “It’s second-nature to glance at the driver of the approaching vehicle before stepping into the road. Understanding how this translates in tomorrow’s more automated world is important.” In what it terms “trust research” at this stage, Bennett’s team is essentially trying to find out of it needs to go to the extent of drive.ai or Apple in so blatantly signaling intention via screens, or whether a simple look of acknowledgement will suffice.
I’m all for this sort of research: it is necessary to establish concepts of how best to connect with other road users. Honestly, I just can’t help but find them a little creepy…
A driverless dark horse?
It may have flown under our radar at the time James Dyson – the man behind the famous vacuum cleaners, Air blade hand driers and many other suckers and blowers – announced that he plans to launch a battery-powered vehicle in 2021, but consider my attention piqued.
Last week, the British billionaire inventor unveiled plans for a 10-mile test track in Wiltshire that the company’s CEO, Jim Rowan, reckons will soon become a "world-class vehicle testing campus". The disused airfield at Hullavington was purchased two years ago and renovated at a cost of £84m. The growing R&D site already houses 400 automotive staff in state-of-the-art hangars and this next phase would take Dyson's total investment to £200m. All with the end-goal of blowing (sorry) Mr Musk out of the water.
The British billionaire inventor recently told GQ magazine that the Dyson car will feature “some” driverless features but that it should not be built like a car with an internal combustion engine. Heck, it may not even look like a car: “radical” is the word he uses.
With this much money and a history of world-class innovation and ingenuity driven by efficiency, I wait with baited breath to see what he comes up with. Sure, the crossover from vacuums to cars is in itself radical – but at the end of the day, creativity will win this race. And I’m pretty sure Dyson’s offering will not suck…
So long, drive safely (until cars are driverless),