Quote:
Originally Posted by Efthreeoh
I checked on the web, Uber costs $1.55 per mile from few websites I saw. The GSA reimbursement rate for private owned vehicle use is $0.55. All Uber is is a non-unionized, independent-driver, Taxi service (with resultant business tax structure) that utilizes a phone app. Uber nothing new relative to the concept of a hired car service and doesn't make the case for AV any more than Taxis do, which have been prior to the automobile. I don;t understand why people try to make an argument that Uber and Lyft are some type of precursor to the automation age; that argument could have been made for the more than the last 100 years. And here's a bit of trivia, out where I live in Central Virginia, we don't have Uber, an Uber driver would starve to death 
So back to the real discussion. My using the air traffic control system was about the prevention of vehicle collisions, rather than deaths per 100 million miles traveled (the main metric used for safety evaluation). 99.999 ground vehicle deaths are due to collisions; a few people do die while driving and then crash. When planes collide in air most everyone dies since the planes fall out of the sky an impact the ground, which is why the FAA tries to prevent air collisions in the first place. I agree, a lot of the reduction in ground vehicle deaths is attributed to safety apparatus and better vehicle design introduced into the automobile since the 1960s, but preventing collisions in the first place is the key to low death rates. Preventing collisions is the primary argument for the move to AV. I'm highly skeptical that AV is actually going to reduce the amount of vehicle collisions while maintaining the same amount of passenger miles driven per year at the same average rate of speed per mile. I certainly do not think vehicle collisions are magically going to decrease while AV is somehow transitioned into the current traffic environment of human-driven cars. I think collisions and deaths will increase during the transition period. Trust me, the DOT is not going to tolerate that situation.
Why I think AV will be less convenient is in the future where people envision a less total vehicle population in the US and a person just summons an AV car service to come pick them up; where AVs are a shared asset, constantly moving about picking people up and dropping them off on individual trips. This will be inconvenient as compared to now, walking out to your car and driving somewhere. It will be inconvenient because the overall speed of travel will be reduced, not increased, so it will take longer to get some place. And as Uber proves, it will cost more per mile. |
So it looks to me you have three points --- it won't work, it will cost more, and it will be less convenient.
As for the first, I guess we'll have to wait and see. I was surprised to read that GM had a petition at the DOT to roll out a 2,500 AV fleet in SF or Scottsdale in 2019. We'll see if DOT approves. GM claims that its AVs are safe.
It's not as if the nascent AV industry hasn't thought about the problems you raised. They're spending billions trying to solve them and the investments are rapidly increasing. The tech guys and the money guys see big progress. All we know for sure is that the technology will improve quickly over time as semiconductor, sensor, and AI improve. I think you agreed that this was feasible in an earlier post. So its really a matter of time. I'm trying to think of another industry of this magnitude where everyone was all in -- in this industry every car manufacturer is devoting massive dollars, the two leading CPU/GPU companies - Intel and Nvidea - are all in, the parts guys are all in, several tech companies (Google) are in, the truck producers are in, and the ride services are all in.
The Uber example kinda disproves your point. Prior to Uber, the cost and inconvenience (as well as regulation) of taxis, limited their use. Today Uber is valued in the billions because it is able to provide on-demand taxi service in minutes with ease of payment. That convenience and cost has caused the demand for those services to explode making it feasible to leave the car at home, do without a car all together, and make a trip that would not have otherwise been made due to inconvenience or cost. The fleet model makes great sense for the urban environment. AV will take over that expanding market when its overall cost is less than an IC Uber. As this scales, there will be less driver-owned vehicles on city streets and less need for street and underground parking. Less congestion. The biggest benefits are in the most densely populated and congested cities. Less congestion means everything is faster. (Unless you assume poor technology.)
As for rural areas, it's likely AV will be individually owned, at least at first due to the lack of density. Pretty simple economics - if the total cost of ownership for AV is less than IC, it will take off. That total cost includes the convenience of not having to drive. It depends how you (and the market) value your time. So when you do your comparisons, make them apples to apples: fleet to fleet or individually owned to individually owned. Also,the convenience factor will expand the ratio of fleet to individual ownership over time. The long-haul trucking future is subject to the same economics.
So at the end of the day, I think your argument boils down to how quickly the technology develops. The cost saving trend is already baked into the underlying declining-cost technologies (batteries, sensors, and semiconductors) and scale..