One of the most challenging parts of writing Ghost Road was dealing with the awful set of terms and classifications we have for AVs. I tossed the Society of Automotive Engineers 6-level scheme for driving automation out the window before even starting. But that left "self-driving", "driverless", and "autonomous" to wrestle to the ground. And as I did, I found each more off-putting and pejorative than the last.
In the end, I didn't come up with anything better to call these new rides. But I think I did sketch out an RFP for a better term. Unlike "horseless carriage" or "cellular phone", engineers' terms that define the new in opposition to the old—we need words that explain the benefits of this new technology to everyday people. Automobile and mobile phone did that, and they stick with us today. It's all there, in Chapter 2, "Deconstructing Driving", if you want to get lexical with me.
The AV industry knows it needs to move on. Musk essentially stole Autopilot from Cadillac, which launched its first cruise control feature under the same name in the late 1950s. GM came upon SuperCruise by, well, putting the word "super" in front of—the AV equivalent of "put a bird on it" I suppose. That kind of stuff won't cut it.
But most of what we're seeing is still pretty measured. Waymo is rebranding everything it does from "self-driving" to "autonomous" to distinguish vehicles that "handle the entire task of driving without depending on a human driver to take control." Today, with no fewer than three well-banked firms rolling out robotaxi services in the West, and the AV industry is still trying to figure out what to call these things. GM's Cruise division still dryly refers to its Origin prototype as a "vehicle", though Motor Trend prefers to call it a "pod car".
One, though, seems to be steering a peculiar and potentially promising course—Zoox, the newest feather in Jeff Bezos' cap after Amazon's billion-dollar acquisition of the cash-strapped startup last summer. Zoox, of all things, wants us to think of its four-seat robotaxi, where you sit facing other passengers, a near-perfect clone of Cruise's own, as a "carriage".
I recognized the imagery instantly. Back in 2016, during my first ride in an EasyMile EZ10 driverless shuttle, I cruised along the downtown Tampa waterfront with delegates to the Florida AV Summit. The appeal of the new mode of travel revealed itself slowly to me:
On the return trip, my attention turned outward. With their wraparound windows and raised coach floor, driverless shuttles provide a perspective on the street akin to what nobles might have experienced traveling by open-air carriage in years past. Unlike the subway, where you're cut off in a high-speed bubble, on the surface the theater of the street surrounds you. Silent, gliding... far more than any experience in a self-driving car, this may be as close to a mythical magic carpet ride as we're likely to get. (p. 60)
Even more intriguing about the Zoox machine is that automation does more than just take the wheel. As Brad Templeton points out in Forbes, "There is 4 wheel independent suspension under computer control, to smooth out the ride." Artificially-intelligent autonomous suspensions, powered by their own computer vision eyeballing the poorly-maintained post-pandemic austerity budget pavement ahead, may be the killer app for future transit. How else will self-driving buses deliver a light rail-grade ride at a fraction of the cost?
It's so important that they do. Because the stars are lining up to create an urban mobility nightmare instead. First, transit systems are in free fall as people opt for private modes of transport. Even the ones that serve big CBDs with good prospects for recovery, like New York City and London, are looking at years of ridership well below pre-pandemic levels. NYC Transit and Transport for London both recently released projections showing 20 precent ridership drops well into the mid-2020s. And that's just for rail. Bus networks are the real victims and ride-hail has picked off riders over the last year who will never return. That's why, unlike almost everything, this is one product where the American version is down-sized rather than super-sized. With just 4 seats, these are not transit vehicles. They're transit-killers.
What's more, they're not some paasport to a future of shared ride-hail mobility, etiher. That future has been vaporized faster than a Texas mask mandate. Ride-hail operators lose more money on shared rides than private ones. The only reason they're still offered is because they serve as a sort of gateway drug for price-conscious first-time customers. With the pandemic, however, that acquisition channel has become less effective. Safe, clean, convenient robotaxis may simply kill it altogether.
And so one altogether plausible future presents itself. We slide into the twentysomething years of the 21st century wrapped in our self-driving bubbles of sanitized steel, slowly creeping along through the mounting traffic. Will we call it a carriage, or a cage? We will have to wait and see.
The urban ushers are here. The line between robots and AVs is already fuzzy and will blur much faster than we think. Law enforcement agencies are eager to push it to their advantage. Police Robots Are Not a Selfie Opportunity, They’re a Privacy Disaster Waiting to Happen (EFF)
Talk about a monkey on your back. Honda R&D's Ropot doesn't move, but that's a small comfort when you consider that the company's engineers are basically proposing to send kids to school with cybernetic familiars to keep them safe as they cross the street. It won't be long before your kid comes home crawling with bots, literally.