There once was a conference (PDF) where experts tried to find a solution to the increasing problem of congestion, pollution and the associated ill-effects of the transportation system. The attendees didn’t find a solution and it looked like modern civilization was headed in the wrong direction.
It was the 1890s and the solution, of course, which the experts didn’t envision, was the relatively rapid rise of the automobile. Speaking at the SV Forum’s “The Future of Personalized Vehicles” event in Menlo Park, CA. Stefan Heck, Ph.D Consulting Professor at Stanford University used this story to compare where we are today with the autonomous automobile and how society is once again set for a major technological disruption.
The disruption will come from a much more efficient utilization of resources, thanks to automation To prove his point, Heck cited some interesting statistics regarding the personal automobile:
- Cars average approximately 4% utilization in the U.S. That is, approximately 96% of the time, cars are parked
- Breaking the above statistic further, according to Heck, only 2.6% of the time is the car serving its goal of moving a person from place-to-place, while 0.8% is spent looking for parking and 0.5% is spent in traffic congestion.
- Further, less than 1% of the gasoline is used to move the people in the car (e.g. most the gasoline is used to propel the car)
- A significant amount of real estate is dedicated to parking and roads in cities (Heck suggested 1/3, which is on the low-end compared to this estimate).
Heck’s answer to what will solve these issues and find a solution to today’s seemingly intractable transportation problems is what he terms ACES – Autonomous, Connected, Electric and Shared. That is, the technologies described by those words are converging and will dramatically change the efficiency and, by extension, the economics of transport.
He suggests that the cost of transport will drop from the 67 cents per mile of today’s individually owned car to an automated ridesharing service that will cost 10 cents per mile. And, as he points out, this approach will be one of universal access, as one does not need a driver’s license (e.g. the blind or tween) to use an automated vehicle.
Jonathan Matus, Founder/CEO of Zendrive, reinforced Heck’s point that ridesharing will be a leading factor for bringing autonomy into the automotive mainstream. He pointed out that we may even see a purchase of an automobile manufacturer by a ridesharing service. He suggested that ridesharing services will have different priorities and will want automobiles that can last a million miles and that the only way they may get this longevity is by owning or, at least controlling, the manufacturing process.. His prediction was the most aggressive among the panelists, as he suggested that we will see mainstream autonomous automobiles via these ridesharing services by the end of this decade.
Although the consensus seemed to be between 5 to 10 years before an individual could purchase an autonomous vehicle, all agreed that regulation and legislative actions, not technology, would be the pacing item for the rollout of autonomous vehicles.
Steve Goldberg of the venture capital firm, Venrock suggested that, “We have already gone a long ways on cost curves for communications, processing and multi-sensor fusion.” He cited his personal experience of working with DSRC (Dedicated Short Range Communications) in 2005 and how that technology is coming to fruition in the nascent Vehicle-to-Vehicle and Vehicle-to-Infrastructure rules being developed by the National Highway Traffic Safety Administration.
Reinforcing Goldberg’s point that the technology is well on its way down the cost curve, Danny Shapiro of Nvidia stated that the fastest supercomputer of 15 years ago required 5k Watts of power and 5k Watts of cooling and an enormous physical footprint, while Nvidia’s recently announced Tegra X1 has the same 1 Teraflop capability, but in a chip-size form factor and only requiring 10 Watts of power (see this example of how laser eyesight was added to a car using Nvidia’s previous generation GPU).
Mark Platshon of BMW iVentures likened the changes we are about to see to what happened with the Gutenberg press. He suggested looking beyond the obvious and seeing the tertiary impacts of adding autonomy to vehicles. His comments captured the essence of the SV Forum’s excellent conference.
Stay tuned for a future interview with Platshon where he will elaborate on his thoughts about the future of autonomous vehicles.