I veiw teslas FSD as not much of a diffetentiator to other products, gm, ford, honda, etc. i worked with automated guided vehicles in warehouses in a past life. The sensors, lidar, etc are all similar. Obviously outside environment is more uncontrolled.
I like adaptive cruise, blindspot warning, lane keeping assist, 360 camera, emergency braking, however, I also got a false alarm driving over a cattle grate out west. Spilled coffee, etc.
I’m a big NO! What happens when it snows? I don’t want to be driving on a 2-lane highway while the fog lines and centerline are covered with snow and ice, and there’s a driverless car coming my way in the opposite direction. No bueno.
It’s fun to imagine what transport would look like in a large city if you were building it de novo with today’s technology and infinite money, like a Saudi Arabia.
I’m thinking a system where everyone could buy or rent a car which, outside the defined city boundaries, could drive at the direction of its driver like a regular car (but could be as autonomous as today’s best self-driving cars as an option).
Within the city though the entire city’s traffic flow could be centrally managed, so you’d type in where you want to go and your “car” would pull you into a flow of traffic, jumping seamlessly from stream to stream in accordance with what real-time data indicated would be the most efficient route. Every car its own carriage in a network of constantly blending and dividing trains. Cars could drive inches apart, they’d never crash, no-one would speed, or run a red - who really likes driving in a city anyway?
That’s how I’d set it up. In any event, it’s weird to watch us use technology to “back into” a system a bit like that. To watch companies like Waymo pour ungodly amounts of computer power into trying to make cars interract with an antiquated infrastructure system. It’s like deciding the only way to stream movies is to have drones fly into your living room every evening with all the servos and cameras necessary to enable them to unbox and insert a VHS.
It’s a big hell no for me, but it does feel inevitable that is how we will all get around sooner than later.
Our city is gung ho on banning private car use. We recently amended our city documents that dictate how the city will be expanded and developed over the next 20 years, and all references to private passenger vehicles have been removed. So, moving forward, all road development will only consider bikes, pedestrians, mass transit and commercial vehicles.
I’m all in. The technology is already better than most drivers in most conditions. But…it’s getting that last 0.01% that is essential. I heard the CEO of Mobileye describe the failure rate goal as equivalent to that of airplane wings on commercial aircraft…and you don’t hear about wings falling off of airplanes very often.
The roads where I live are too narrow and busy for cycling most places. I see autonomous driving as a fix for this. It is coming in stages…features that prevent cars from hitting things will continue to be added and eventually standard. Level 5 autonomy (no steering wheel) has been doable for several years as long as you allow $100K worth of computing power and sensors and you only operate on finely mapped roads and good conditions. The compute costs will come down as the algorithms improve and solidify, and sensors can come down as they go mass market as well. As vehicles become more and more “fly by wire” they will more and more prevent you from doing bad things and then eventually we get full autonomy.
Another way I like to think about this is comparing decision-making speed, e.g. do I steer into the crowd of kids to avoid a head on collision. For a human you have to make that decision in a fraction of a second. An autonomous vehicle has the decision pre-made and executes it in microseconds.
I like the driver assist features for the most part. And I think that is probably going to continue to grow.
But autonomous driving fails in the situations that are the most difficult. Snow, rain, fog, bad roads, etc. And in those situations it is going to hand control back to drivers who will have less and less experience actually driving.
I don’t know if we will get there in moat situations in the next 30 years. The infrastructure improvements would be enormous.
We always get these fanciful predictions about how the technology is going to take over almost immediately. It just isn’t so. In 2015 my employer, Mega Insurance Corp. told us quite definitively that in 10 years we would not be writing car insurance because the cars would be self-driving. It is 10 years later and I’m not sure we are actually closer to that being reality than it was in 2015. If anything the momentum seems to have gone the other way.
I’m all for it, so long as it is a voluntary choice. Most drivers suck, so I’m a fan of getting them from being in control of a car. When I walk my son to school, I get to see so many people on their phones driving through a neighborhood, with kids around.
Now, will there be stupid accidents and some rare fatalities with autonomous vehicles, I’m certain of it. However, from a macro level, I would expect it to be better compared to today’s current crop of distracted drivers.
Also, when I’m in the Midwest, maybe I’ll encounter less FIBS, and drunk Sconnies.
Having once worked in that space, a huge fan, when done correctly and explained to folks, Waymo and Cruise were doing it right. Tesla’s move into the space, the reduction of requirement’s by NHTSA. will make it a shit show.
Then again, most every other aspect of life in America is becoming a shit show, so why should our roads not go back to the 1960’s like everything else.
Given that they have been operating in San Francisco for a number of years, cloudy/overcast weather is not the problem. Rain and snow are possible challenges, but they have had to deal with the former in SF as well.