Autonomous Vehicle Technology:
Where are we,
Where are we going,
How to get there
Prepared remarks made at
Unmanned Systems Caucus
Co-Chairs, Reps. Daniel Lipinski (D-IL), Joe Heck (R-NV)
Wednesday, July 8th, 2015
Briefing: 1:00 – 2:00pm
2325 Rayburn House Office Building
by
Prof. Alain L. Kornhauser, Ph.D. F. ITE
In the first half of this year, the perilous safety of our surface transportation systems has been much in the news and close to my home. The Amtrak derailment[1] near Philadelphia, the loss of Alicia and John Nash[2] on the New Jersey Turnpike and the totaling of her “Beamer” by my friend’s mother have made clear to me how ineffective we remain in dealing with the safety of our surface transportation system.
Existing technology could have inexpensively and easily averted each of these tragedies. Safety, the fundamental reason why there is public oversight on transportation, remains THE problem.
What has been happening?
One thing is becoming clear. Technology may be able to help because it is becoming more reliable than humans.
If we look at the reliability of humans versus automation in the driving function, we know that in the past, yesterday, humans, while not perfect, have been very good and unquestionably better than automation. Today, unfortunately, humans have gotten worse and maybe substantially worse. The culprit has been connected distraction; “Texting” or whatever we want to call it.
In the mean time, Automation, through “Moore’s Law”, has been making substantial improvement.
Tomorrow, the expectation is that automation will continue to improve substantially while there doesn’t seem to be a cure in the works on the human side. In fact, we may have already, or are soon to, reach a cross over at point where we would be better off to have automation drive at least some, if not most, of the time.
The issue may be well be “will Automation Sell”? We know that to date, Safety hasn’t sold, will automation Sell??
Why hasn’t safety sold? Why have we had to have “Click-it or Ticket” campaigns to get people to wear seat belts anyway? Why doesn’t everyone drive a Volvo? Probably because we each feel that we are safe drivers. It’s the other guy!
Plus our focus, rightfully so, has been on crash mitigation. Seat belts, air bags, crush zones. We’ve done a great job in Crash Mitigation. Lives are saved; Injuries are less severe; however, the cost to fix mitigated crashes has gone up.
In a recent report, NHTSA put the economic costs of roadway crashes at $871 Billion per year in the US. Of that 2/3 are societal costs associated with the personal traumas and pain that crashes impart on individuals.
The remaining 1/3, or about $280 billion, are auditable dollars that go directly to fix things. We as individuals “self-insure”, or “eat”, about half of that through insurance deductibles and taxes that are subsequently spent by local, state and federal government agencies on our behalf.
But, the other half is meticulously paid by Insurance. These are real auditable dollars that are monitored very closely. These go to lawyers, undertakers, MRIs machines, body shops etc. and amount to about $140 billion per year. These are expenses that are paid after we’ve done all the good work in crash mitigation. These are the auditable liability Losses incurred by the insurance industry.
However, if crashes didn’t happen, these monies would not be spent and would appear on insurance balance sheets for everyone to see.
While avoiding all crashes is not realistic, half may well be achievable. This would free up about $70 billion per year to pay for the crash avoidance technology that delivered those dollars.
More importantly, if looked at this from the perspective of the individual vehicle we may well be approaching a situation in which:
The amount of real cash that is saved by insurance may well be greater than the annualized cost of the crash avoidance technology that delivered those savings.
Present Value {Expected (Loss Liability w/o the Technology - Loss Liability with the Technology)}
Cost {Automated Collision Avoidance Technology}
Once we hit this crossover point where the quality of the collision avoidance technology is so good and its cost is so inexpensive that insurance makes money even if they have to pay for the technology, then this is a real business case. Safety may finally sell and it is insurance that is buying because they make more money. They would be stupid if they didn’t.
We may well have reached this point with the Transit Industry where there is a mature system monitoring liability expenses associated with bus crashes.
Over the past 10 years more than 1,000 people have been killed and more than 150,000 injured in crashes involving transit buses
More importantly the industry has spent almost $5 Billion in reported liability expense, much of which came from the top line of a heavily subsidized, largely self-insured, industry.
To put it more bluntly, in the next 5 days the industry can be expected to spend $6.8 million, cash, to cover bus crashes.
And, many avoidable expenses aren’t covered by the reporting, so the actual expenditures could be substantially more.
Looking at 2013, nearly a half billion was reported to be spent or over $6,000 per bus per year.
Automated Collision Avoidance technology can be expected to cut that number in half, making $3,000 available each year over the more than 10 year life of a bus to pay for the technology. This is a very attractive Return-on-Investment, RoI, to each transit agency. Plus all of the societal benefits of live saved, injuries averted, come along for free.
A similar situation exists in the motor carrier industry, which is substantially self-insured, and the RoIs are even more attractive.
What is also important is that this automation is focused on improving the work place of those that are required to drive as part of their employment, the bus driver and the truck driver. Improvement of workplace safety should be supported not only by insurance but also by organized labor and OSHA.
If drivers were miners, or assembly-line workers, OSHA would be there.
Where can this go?
Taking the focus of automation as the means of crash avoidance implies that “Moore’s Law” technology may well apply. That is, technology that both improves and becomes cheaper over time. That improvement will likely lead to automation that is able to serve the complete trip from origin to destination.
Today elevators, airport people movers, some metros are completely automated. Some vehicles are beginning to operate in gated communities such as LaRochelle, France and maybe soon, Moffet Field and Fort Momouth.
As the technology improves I can easily envision NJ Transit managing 1.5 million shared autonomous taxis to provide mobility safe, clean and energy efficient mobility to more than 80% of the 30 million trips that take place in New jersey on a typical day. Quite a change from the less than 4% that they serve today. A real revolution in public transit!
How do we get there?
Easy…
1. Nurture the environment that will create the automation technology that will follow “Moore’s Law” … Get better and cheaper over time.
a. The private sector is already there: Google, Mercedez, Ford, Subaru, Volvo, Elon Musk, MobilEye, Roscoe, Soterea, etc. Allow and enable them to flourish.
2. The public sector is way behind.
a. 18 months ago I submitted a proposal to Federal Transit Administration to develop the necessary standards that would allow the transit industry to begin to readily procure crash avoidance technology. The proposal was supported by a major transit agency, the American Public Transit Association and major Transit Insurance entities.
It was rejected. Thus, the whole process of acquiring the technology has been set back at least 18 month.
The eventual cost to the industry of an 18 month delay in starting on this safety road is at least $350 million top line, 75 lives and 100,000 injuries.
b. DoT is about to spend some $500 million of a new round of “Tiger” grants, none of which are focused on safety-related automation.
c. On the other hand: I’d like to applaud Congressman Lipinski’s legislative intuitive to create a Future Transportation Research and Innovation for Prosperity (TRIP) Act to support research, development, certification and commercialization of the technology that will allow us to achieve a substantial improvement in safety and mobility.
3. I’d like to end by presenting a short video on one of the things that I am funding to help accelerate the whole process.
https://www.youtube.com/watch?feature=player_detailpage&v=yARbNYcjPQM
Alain L. Kornhauser
[1] http://www.nytimes.com/2015/05/14/us/amtrak-train-derails-crash-philadelphia.html
[2] http://www.nytimes.com/2015/05/25/science/john-nash-a-beautiful-mind-subject-and-nobel-winner-dies-at-86.html?_r=0