Contrarian Thinking on Autonomous Cars

I live in the heart of car culture, Los Angeles, so I spend a lot of time behind the wheel, mostly listening to podcasts.

This week I listened to a great @a16z discussion on how autonomous vehicles will change cities and transportation. About three minutes in, as I zipped into the FasTrak lane, the conversation turned to how autonomous cars will free up large amounts of space in cities — parking lots, car repair shops, parking spaces — all the infrastructure wasted on machines that spend most of their useful lives idle.

This makes tremendous sense … at first. But what if the opposite is true, that self-driving cars increase inner-city congestion? Sure, parking garages won’t need to house cars sitting idle all day, but where will cars sit while they wait to be summoned? Will they drive around aimlessly? Won’t they need to be “held” somewhere until they’re needed?

Fair enough, enthusiasts may say, but look at the math. The number of trips people take during the day, once they’re at work, is so much less than the number of cars now parked in lots. Surely it’s a net gain?

Well, let’s think about that. If people turn their commute into an extension of work, and if they are somewhat indifferent to the time they have to be in transit, won’t more people want to take a car into town each day? Those cars will have to go somewhere and if demand for short haul “downtown” trips are much less than long-haul work to home trips, some of the latter will have to be stashed somewhere during the day. Maybe they drive to a holding lot on the outskirts of town until around 4 p.m. when they can head back into town to pick up their renter du jour. It may be that self-driving proponents are right, software will make it all better and more efficient, but maybe not.

OK, now let’s think about that car as an office. Let’s see, I’ve got an HR call at 7:30, I need to talk to Jane about Jack at 8:00, and then, oh jeez, I have to get that TPS report done by 9. No one wants to listen to me yammer with workmates, and I can’t abide someone yammering while I’m trying to finish a document, so I guess I will travel in alone. Times a million.

Sure, most commuters already drive alone. But some people who now take mass transit — especially expensive mass transit in cities like Washington DC — may decide commuting by self-driving car is better than being sardined in a sweltering Metro car. And so the volume of suburban to inner-city drivers will grow, just as gentrification encourages downtown-living young professionals to order service from their apartment to the office. We may want self-driving cars to reduce congestion, but wishing it will not make it so.

(Side note: one clear winner if self-driving cars take over the daily commute will be mobile broadband providers. Extending the workday while traveling will drastically increase demand for high throughput mobile broadband. Employers will soon see demand for a new job benefit — reimbursement for “mobile transit” data use. Perhaps fleet managers will compete on the quality of their broadband service or built-in apps to conserve data use.)

But let’s get back for a minute to that fast-changing inner-city landscape. With thousands of people zipping into work in cars they don’t have to park, what does drop-off and pick-up look like? Anyone who’s ever managed the drop-off line at a K-12 school knows how congested and contentious the simple act of coming to a stop, disgorging passengers, collecting belongings — and oh, sorry just need to grab that thing I forgot on the back seat — thanks everyone, then quickly back to the trunk, then dash inside. It’s a mess. And on a two lane downtown street, with ten cars lined up on both sides of an intersection, even with a prohibition on street parking?

Another possibility is that the other great Silicon Valley dream, VR, advances as quickly as self-driving cars, offering a substitute for driving altogether. Communal work spaces located in suburban office parks with phone-booth style VR meeting spaces — or even stripped down home office versions — will enable people to skip the long commute.

So the prospect of self-driving cars for commuters is very enticing but very complex. Why not begin a broader rollout with services for retirees or the elderly, and teens — customer segments where demand, logistics, and regulatory acceptance more easily coincide? Both need what self-driving cars aspire to provide — safety (in terms of getting where you want to go without having an accident) and reliability, on-demand.

An increasing proportion of retirees and the elderly dispersed throughout cities and the suburbs are “aging in place.” A key contributor to healthy aging is socialization, and the biggest impediment to meeting up with others is the difficulty or cost to drive. That suggests strong demand for a service that caters to transporting retired or semi-retired people to gathering places, to the movies, a part-time job, the country club, the coffee shop, or the mall. This is a service many would pay for — end users, their adult children, even auto insurers might subsidize it, if it means lowering the risk profile of their insureds. Self-driving cars will build political and commercial suppot most quickly if they increase efficiency or quality of life without noticeably worsening existing congestion.

The same is true for transporting teens to after-school activities, friends’ houses, to night time activities, or back home. Many parents may find it more comforting to have their children transported by a bot than a human, provided they have real-time video/contact with their children, and always-on monitoring to verify pickup, current location, and delivery. (And yes, any system that transports children will need to have multiple security layers to stop hacker/kidnappers from stealing kids or making cars crash.) Self-driving teen transport services might be the final nail in the coffin of a nightly family dinner, but it will also be a liberation of sorts for parents frazzled by efforts to balance work and carpool commitments.

Nothing would be finer than a commute that freed drivers from driving. But big cities are complex. Infrastructure built to a certain purpose is hard to unbuild. Products that we hope to lessen congestion may actually increase it, along with pollution. Transforming the commute is a laudable goal but extending safer mobility to those who need it most may be a surer, if more circuitous, path to that larger goal.