Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla CEO Musk on Battery Factory, Model X, Apple, autonomous cars

This site may earn commission on affiliate links.
Great video. I was a little dissapointed that he didn't comment stronger about iOS or Android integration when asked the question. I was hoping they would have pulled a fast one and done something for integrating them into the upcoming V6 software. Oh well
 
IOS or Android is nothing more than a distraction. Embedded Linux is working just fine.

Something else really struck me.

Google has cars out on the road driving themselves.
Elon says Tesla has the strongest atonomus vehicle (auto pilot in T terms) team of any AUTO MFG and MAYBE ANY company. Normally Elon does not say "full of ****" stuff which means there is a tremendous amount of auto pilot work going on at Tesla. He also suggested a few year time frame which would equate to four real years (one Elon development year tends o be two calendar years).
 
He said some interesting things in an documentary here:

Tegenlicht - Hoe raken wij de auto kwijt

I hope it's available outside of the Netherlands. The video is probably 50% Dutch, 50% English.

They are developing an "autopilot" kind of feature.
As going from a car that is 99% safe, to 99.9 % safe is a factor 10, and going from 99.999 is probably still not safe enough. (at 28:00 minutes)
 
This is an area where our current PI environment and lack of personal responsibility collides directly with common sense.

Statistically speaking, it would be possible to produce a system that is safer than a human driving. The problem would be that it would not be perfect. The very first accident, especially one involving injury, will have every PI attorney lobbying all parties to seek damages. All this even though it would be less safe to have humans driving.

As with so many other examples, it will be the financial interests of a few that hold back progress for us all. We have simply lost the ability to properly evaluate risk on a personal level and take responsibility for the decisions we make.

Ayn Rand is smiling somewhere.
 
This is an area where our current PI environment and lack of personal responsibility collides directly with common sense.

Statistically speaking, it would be possible to produce a system that is safer than a human driving. The problem would be that it would not be perfect. The very first accident, especially one involving injury, will have every PI attorney lobbying all parties to seek damages. All this even though it would be less safe to have humans driving.

As with so many other examples, it will be the financial interests of a few that hold back progress for us all. We have simply lost the ability to properly evaluate risk on a personal level and take responsibility for the decisions we make.

Ayn Rand is smiling somewhere.

It should be possible to keep the liability with the driver, then it will be up to each individual to what extent they trust the autopilot, and while some might not adopt as quickly after signing a massive legal disclaimer, enough will pick it up that it will be safer on the roads and the fear about the technology will start to subside.
 
Autonomous emergency braking is a feature that is becoming adopted. These systems pre-charge brakes, increase braking force, or automatically apply brakes when sensors detect an imminent collision. But its deployment was significantly delayed out of fears that such a feature implies liability for the car manufacturer in case of an accident (whether or not the system functioned properly). It is not clear that you can keep the liability with the driver if the manufacturer is offering systems that alter the probability or outcomes of an accident, even if those systems are wholly beneficial.
 
There are very compelling arguments to be made for the above described technologies. I think Elon realizes that legislative protection for manufacturers of these systems is not likely, or likely heavily watered down if actually passed. The logical solution is to "aid" the driver with an auto pilot while leaving the principal responsibility in the hands of the driver (as Martini has suggested).

Again, a logical solution after acknowledging the lay of the land.
 
Remember something else. The USofA not only is not the only country on earth, but we have heard that even in the short-term it is likely to account for less than one-third of Tesla's sales. Most other nations have not been subject to so much destruction from an out of control PI environment as has the US - as a company, Tesla could do very well indeed, thank you very much, by incorporating whatever autonomous driving package in its cars sold to the appropriate countries - I'm thinking China, above all - while our own nation continues to languish ever farther behind as we wallow in our own self-created destruction.


And how's that​ for a cheery end-of-week post?
 
I would think that the first autonomous feature deployed as a step function of where we are at currently would be a platooning capability. You already have DRCC that keeps you a certain distance from other cars. And a similar package will also automatically brake if you get too close to the car in front. It isn't such a great leap to have a platoon of say 5-10 cars driving themselves very close together under the control of the first car in the lead. Since aerodynamic resistance is one of the main causes of poor mileage, this capability would be one of the best examples I can think of for a "low hanging fruit" ready to be picked.

Five cars on the way to Vegas, all driving very close together, knowing they are essentially a single unit. Someone wants to join the platoon they join at the end. Someone in the middle wants to leave, the car is given a command and pulls out into the adjoining lane, then goes manual. The remaining cars then close in to fill the newly created gap. You don't need to have all cars on the road with this capability, but you need that capability to join a platoon. Anybody know what the gas/kWh savings would be from doing this?

RT
 
I would think that the first autonomous feature deployed as a step function of where we are at currently would be a platooning capability. You already have DRCC that keeps you a certain distance from other cars. And a similar package will also automatically brake if you get too close to the car in front. It isn't such a great leap to have a platoon of say 5-10 cars driving themselves very close together under the control of the first car in the lead. Since aerodynamic resistance is one of the main causes of poor mileage, this capability would be one of the best examples I can think of for a "low hanging fruit" ready to be picked.

Five cars on the way to Vegas, all driving very close together, knowing they are essentially a single unit. Someone wants to join the platoon they join at the end. Someone in the middle wants to leave, the car is given a command and pulls out into the adjoining lane, then goes manual. The remaining cars then close in to fill the newly created gap. You don't need to have all cars on the road with this capability, but you need that capability to join a platoon. Anybody know what the gas/kWh savings would be from doing this?

RT

The system would need to handle failure of a car without having a pile-up. I would exoect autonomous systens to _increase_ distance between cars, lowering speed at volume, which would be made up for by smoothing traffic and having fewer delays. Although, autonomous system would allow more efficient rental systems, and allow better ride sharing, so decrease ownership and traffic anyway.