Throughout this summer a small white car with odd instruments bolted to its roof has cruised through the streets of the inner Mission, making frequent, dream-like appearances throughout the day, and leaving puzzled residents behind in its stately wake. Bart Dority, who lives on 25th Street, has watched the car drive past his house for weeks now.”I thought: ‘Who is that in there? Scully and Mulder?’”
I started calling the car “Moby Dick” after the titular whale of Melville’s masterpiece and like Ahab finally pursued the whale after being provoked by the sixth sighting of the day while chatting with my neighbor. “What is that?” my neighbor exclaimed in her lilting Russian accent. “I see it all the time! But I don’t know what it is!”
I ran down the street and caught up with the car. Three men sat inside staring at computers in their laps – all of the Mission cars, it should be noted, still have humans inside. The men looked up at me warily.
“What kind of car is this?” I asked.
“We can’t tell you,” the man in the front passenger seat replied.
When pressed, he shrugged. “We can’t tell you,” he repeated. He seemed happy with his and the car’s anonymity.
“Look up git cruise,” another man, sitting in the back offered.
“Git cruise?” I repeated, unsure I’d heard him correctly. “Yeah,” he replied. They drove off.
They didn’t have to bother engaging in subterfuge: their privacy was stripped away in May when The Verge reported that driverless cars or AVs — autonomous vehicles — were being test driven in San Francisco by a former start-up called Cruise Automation, started by a man named Kyle Vogt who developed technology that will allow any car to be fitted with a self-driving system. Cruise Automation started test-driving the first car — described by one of Vogt’s collaborators as being “pretty ghetto” — on the 101 freeway sometime in 2014.
The car had its first accident at Bryant and 7th streets on January 8th, according to the accident report filed with the California Department of Motor Vehicles . The Chevy Bolt AV was traveling about 20 mph when it began to drift back and forth between the right and left lanes on Bryant Street. When the driver switched the car back to non-autonomous control, he forgot to correct the car’s trajectory, which crashed into a parked Prius. Both cars suffered “minor visible” damage. “There were no injuries,” notes the accident report, presumably referring to the rattled human non-driver.
Safety, or course, is the big question for driverless cars and their makers: unlike the sleek AV seen driving smoothly through in this Mercedes commercial, real AVs must contend with the chaos of urban traffic and all the elements therein. Happily, California requires AV manufacturers to submit collision reports, allowing the public to assess the relative safety of the cars for themselves.
There have been 16 reported collisions since 2014, most of them involving Google’s AV and most caused by human error, according to the filed reports. Only the accident involving Cruise Automation was caused by a system malfunction. In anticipation of more accidents happening as testing in cities increases, the Department of Transportation has committed $4 billion dollars to assist the manufacturers of AV vehicles and technology to improve the safety of their cars and “accelerate” the development of AVs. Cruise was acquired by GM for $1 billion in March, and will undoubtedly benefit from some of the DOT largesse. In contrast, the DOT is spending $1 billion on their Clean Communities Grant Program, a “competitive grants” program designed to encourage city and regional governments to improve transit, pedestrian and bicycle transportation options.
As of August 17, Cruise is slightly behind the curve in the rush to get AV’s on city streets: This month, Uber, the car service formally operated by humans, will introduce its first self-driving car to its fleet in Pittsburgh. Judging from the number of test drives in the Mission — there are at least seven a day, based on eyewitness accounts — Cruise can’t be far behind.
When will Cruise unveil its technology in San Francisco? Who knows? Cruise, like its engineers, isn’t talking. The secretive staff are supported in their reticence by an equally secretive website, which, aside from acknowledging their frequent test drives on the streets of the Mission, is free of any other information; the site is whited-out, figuratively speaking. Silence seems to be a company tradition: Vogt declined to be interviewed for a story back in March, and neither he nor anyone at Cruise responded to my questions by press time.
The benefit that driverless cars bring to an increasingly crowded urban environments is debatable. Manufacturers claim that driverless cars use parking space more efficiently and will be a boon to disabled or elderly populations. From the perspective of pedestrian and bicycle transportation advocates, the benefits are dubious. “We think that without public ownership of AV fleets, the use of driverless cars will exacerbate the privatization of the transit system and that this technology will increase private vehicle use,” Dave Snyder, Executive Director of the California Bicycle Coalition told me. “From the point of view of social justice, a more important investment than autonomous vehicles is an investment in protected bike lanes to support bicycling, which happens to be the most equitable and affordable form of transportation.”
San Francisco’s Transit First policy wants to reduce the number of single-occupant automobiles in the city. But technology which transforms a privately-owned car into something like a personal BART system might have the opposite effect.
The promise of release from the labor of driving might also prove illusionary: the makers of these systems point to the benefit of freed-up time to get stuff done, making the driverless car sound more like a mobile office.
Moreover, while our streets are public, this technology isn’t. Since Cruise won’t talk, it’s impossible to know if their wizardry will benefit public transit. Driverless cars may look like the future, but all things considered, they may simply be more of the same: the same amount of — or even more — cars on the street.
I can see a situation where someone will have their car drop them off at work, then flip on their uber app and let the card pick up extra income by giving people ride while the owner is working. At the end of the day it picks the owner up.
The flip side of this is more people will have access that car during the day without needing one of their own. Perhaps reducing the overall need for ownership.
That combined with new legislation that allows people to convert garages to housing units may mean More housing and less cars.
So, professional drivers are paying taxes that are spent specifically to put them out of work. The rich get richer, there is no further thought put in to the automation agenda. How/whether drone cars can lead to safer, cleaner, better anything is hazy at best. And with zero legal precedent, good luck figuring out who in the full stack of software/hardware/holding-companies you can sue when one of these contraptions runs you over in the crosswalk.
Safety is one issue, an important one. Congestion is another. How does this not add to the already congested SF streets? A boon for disabled and elderly? Only because our public transit system is weak and getting weaker. The billions invested in driverless cars to boost Uber and friends would be much better spent on expanding and modernizing our public transit system which would have fewer safety issues, reduce congestion and serve not only the disabled etc. but all San Franciscans.
Why all the negative speculation and spin ML?
AV’s will be safer than having human drivers — we humans are the best there is right now, but we are not good drivers — witness all the accidents — we are distractible, emotional, tired, and have many other demands on our attention.
Our streets will only truly be safe and pleasant once humans are no longer allowed to drive vehicles on public streets. (you can always drive cars as a hobby on private land, like riding a horse).