Self Driving Cars Will Save and Improve Lives

If self-driving cars can become ten times safer than human-driven cars and became standard for the world then 1.1 million of the over 1.2 million global traffic deaths would be prevented each year. Self-driving cars would also enable old people and children to get around safely and conveniently with more independence.

Tesla cars are already rated as the safest cars by the NHTSA. The autopilot driver-assist system is making Tesla driving twice as safe.

There are over 100 traffic deaths every day in the USA. Traffic in the USA is many times safer than driving in developing countries like India, China, Africa and South America.

The US population is also aging. We all know of increased accidents from older drivers who should no longer be driving.

We also know that impaired driving is common. People who are drunk or high cause more accidents and more traffic deaths.

NHTSA estimates that 8,730 people died in motor vehicle traffic crashes in the first three months of 2021, a 10.5% increase from the 7,900 fatalities the agency projected for the first quarter of 2020.

These increases in fatalities come even as driving declined; preliminary data reported by the Federal Highway Administration show that vehicle miles traveled (VMT) in the first three months of 2021 decreased by 2.1%, or about 14.9 billion miles. The fatality rates per 100 million VMT for the first quarter of 2021 increased to 1.26 fatalities per 100 million VMT, up from the projected rate of 1.12 fatalities in the same time last year. Non-fatal car crashes are 200 times more common.

In the 1st quarter,

Tesla in Sept, 2021:
Estimated Autopilot miles to-date: about 10 billion miles
Estimated miles in all Tesla vehicles: about 40 billion miles

If Tesla cars were as dangerous as average cars in the USA then the 1.1 to 1.26 fatalities per 100 million VMT would mean an expected 440 to 500 deaths. Tesla cars in China are on roads and with drivers that are five times more dangerous. China has about one-third the numbers of cars per person compared to the US. China has about 260,000 traffic deaths each year. The traffic fatality rate in China is probably about four to seven times worse than it is in the USA. The WHO had statistics that over 700 people per day die in traffic related accidents. Most people who die in China are pedestrians hit by cars. Developing countries in general more dangerous roads. There are more new drivers and the roads and driving is generally less safe. China has about 5 fatalities per 100 million VMT.

There is a site that tries to track all deaths related to Tesla vehicles. The site claims about 210 deaths related to Tesla vehicles. They have statistics of one death per 240 million miles. The site is also tracking deaths in China and other countries. Tesla has about three to five times lower deaths than expected for average vehicle miles driven in different countries.

Ford is recalling about 16,430 vehicles manufactured from January 2 through May 27, 2021. A misrouted seat belt may not adequately restrain an occupant in a crash, increasing the risk of injury. If the driver-side seat belt fails inspection, owners should not drive their vehicle until the repair is performed.

8 thoughts on “Self Driving Cars Will Save and Improve Lives”

  1. Dude, this post is riddled with basic statistical errors. Comparing Tesla autopilot to all miles driven is nonsense since most crashes happen off the major highways which of course is where autopilot is mainly used. There's also nothing to guarantee that a 10x drop in crashes would result in a 10x drop in deaths; if the hardest crashes for a computer to avoid are also typically the deadliest, we might see little if any drop in fatalities. I could go on but you get the point.

  2. That picture of Prince Philip made me think about the edge case of rural/offroad conditions that elderly might want to go to with a robotaxi. That might still take some AI development to do, or shifting to a a palanquin with robot horses as different vehicle platform.

  3. The medical profession is sincerely concerned about this. Not because they want the traffic deaths to continues, but because the vast majority of healthy transplant organs come from vehicle accidents.

    Growing and printing new organs isn't quite ready yet. Neither are permanent prosthetic organs. Nor is regenerating existing organs in situ. Someday, probably, but likely not before driverless tech takes off.

    Could this bring about Larry Niven's organlegging as an actual thing?

  4. You can't measure safety just by miles driven. If you did that for everything, the Apollo space program would look safer than the commercial airline field, measured by miles traveled including to the Moon and number of deaths/mile=0 (if you discount the 3 who died during a test of Apollo 1). The only significant accident was Apollo 13 and no one died there.
    Tesla has picked the low-hanging fruit of mostly highway driving and some suburban driving, mostly in sunny areas during the day. It's still not FSD and is human-assisted on the critical but significant left turns and dense urban areas.
    In the third world, street lines are often absent even on 2-way streets, people wander freely all over highways (I was in India and was shocked to see people crossing 7-lane (minus lines) highways en masse. no one was hit and apparently this is considered normal). People ride 3-4 (with kids) on motorcycles designed for 2.
    Microcars are common everywhere but the U.S. They are safer to drive, but more deadly if you still do get in an accident.
    Tesla's stats are rigged by favorable VMT situations. FSD is less predictable and robots get in accidents in situations humans are good at, while being good at situations humans are bad at – long, monotonous sleep-inducing trips.

  5. I take care of my 94 year old mother who just stopped driving herself and really misses the freedom. Self driving cars would make a big difference to the independence of many elderly people.

  6. Perhaps in an emergency, vehicles could be operated by gaming controllers. Works fine for driving simulators.

  7. After all, we can't all afford chauffeurs. Few consider how hard it is to live without being able to drive in most places, until they have to do it. In rural areas, unless someone "services" you, it's impossible.

    My beloved grandmother had to leave the farm house she, and my grandfather built after living there 50 odd years, because she could no longer safely drive. Thankfully, she could move into an apartment in our house, where she could have her own place, yet still be with family.

  8. Logic is on your side (about replacing maually controlled cars with self-driving) but I would probably miss driving my car myself. It might also make the long anticipated "air-car" possible; if the "pilot" is an AI. Of course once the changeover happens said system would be a tempting target for saboteurs/hackers/terrorists; imagine being able to crash the entire US transportation system in one fell swoop. Sure the "experts" would assure us such a thing would be "impossible" because of all the safeguards.

Comments are closed.