Q & A with Autonomous Expert, Transportation Professor Kara Kockelman
I recently chatted with University of Texas professor of transportation engineering, Dr. Kara Kockelman. She’s conducted extensive research into self-driving, automated and connected vehicles; travel demand and safety; roadway pricing and emissions; and other topics.
What were your biggest surprises from the recent conferences on autonomous vehicles?
One statistic that stuck out was a labor economist noting the worst-case added unemployment we can expect in coming years, due to automation of travel, would [only] be +0.5% (half a percentage point), and would be anticipated only if during a recession.
Keeping emissions down will help avoid the approx 50,000(!) early deaths (of older people, with compromised heart & lung function, due to long-term pollution exposure from roadways) we have each year in the us (& perhaps 1 million world-wide) from such emissions.
To what extent would you say that, on balance, you are optimistic about how driverless technology will affect society?
I think it frees a lot of drivers’ time, but not their passengers’ time. It allows new people to travel more freely, it allows many people to let go of a car, and hopefully never experience a fatal crash or know someone who does. But it can also encourage people to stop walking and biking– even if just walking to pick up one’s car in a parking lot a block away. It’s not healthy in this regard. It can gridlock our roadways and speed up climate change [with the increase in total travel]. Cities need flexibility in managing vehicle automation, and smaller vehicles, fleet-managed, shared, autonomous vehicles need to be key modes. Credit-based congestion pricing is also key.
Tell us about customer demand for driverless services once they have matured.
Drivers will really appreciate the opportunity to read, sleep, and converse with other passengers. It’s a “killer app” in this regard. But the crash benefits, once monetized, are roughly as valuable to our society. As consumers notice their neighbors and friends staying safe on all sorts of trips, see cost savings and travel flexibility, such vehicles will become very common.
We hope that use of shared AVs takes hold of Americans– who are hesitant to share rides with strangers, and ride-sharing/real-time carpooling services can be coordinated in efficient vehicles– before people are offered a chance to buy their own. We need a shift in behavior to avoid further harm to humans and our planet. Paying by the mile (vs paying a lot of money up front, as a sunk vehicle-acquisition cost), helps people avoid over-consumption. Cars and trucks come with a lot of external costs, and Americans already over-consume scarce road space, air quality, and energy.
According to Sebastian Thrun, “a mile in a self-driving robotaxi will cost about $0.30/mile, which would save families a lot of money.” Would you bet on this cheap future for transportation? Do you have other cost estimates?
$0.30/mile is feasible in the long-term, in smaller vehicles. Nearer-term, I imagine $1/mile, which gives 20% return on investment to the fleet provider, beats taxis, ride-hailing services, and helps a lot of travelers keep costs down, but not everyone.
How should cities be approaching transportation and infrastructure spending and planning? (Nashville and other cities have recently nixed infrastructure spending, often citing uncertain future of transportation.)
Road-miles are difficult to add in well-developed locations, and we know we can’t build our way out of congestion, thanks to pent-up demand. It’d be a lot easier to apply credit-based congestion pricing and manage our systems intelligently. Higher gas taxes are needed regardless, in most states, to reflect true cost of our travel, but EVs don’t pay [gas tax], so vehicle miles traveled fees are needed to maintain existing systems.
You’ve mentioned airline concerns. I’ve research this as well. I envision ordering a ‘sleeper’ AV from Austin at 11pm and waking up in New Orleans at 7am– cheaper and more convenient, having slept the whole time. No time wasted. No hassle in an airport. Is that feasible? Would you be worried if you were Southwest Airlines?
Yes, I would expect lower demand eventually, but distances greater than 500 miles will still favor air travel.
What is the single biggest misconception about driverless transportation?
The biggest misconception is that they’ll solve our transportation issues. Congestion, energy, emissions, climate change, human health, obesity become bigger issues. Crashes could also rise if people abuse the technology (e.g. pedestrians playing chicken with cars). We need small, shared, efficient vehicles.
Have you looked at individual company business models? Tesla? Waymo? Company-owned fleets of robocars vs individually-owned-and-lent robocars vs gov’t-operated fleets?
I’m not sure which way they’ll go. They don’t really know either, but partnering is going to be key. Enterprise, Hertz rental cars, etc. are also players here.