Facts don't care about your feelings. Tesla's cars have had a much reduced fatality rate. When they are ultimately allowed to be on the roads without drivers, they won't be perfect, but they will save millions of lives as they will never drive drunk, be distracted, have medical problems, drive negligently or recklessly, or simply be incompetent.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple Reportedly in Process of Acquiring Self-Driving Vehicle Company Drive.ai
- Thread starter MacRumors
- Start date
- Sort by reaction score
Facts don't care about your feelings. Tesla's cars have had a much reduced fatality rate. When they are ultimately allowed to be on the roads without drivers, they won't be perfect, but they will save millions of lives as they will never drive drunk, be distracted, have medical problems, drive negligently or recklessly, or simply be incompetent.
I can’t imagine Apple ever being interested in Tesla, despite the recent rumor. The last thing they’d want is a train wreck of a manufacturing line. I’m sure Apple’s already put far more resources into the elements they’re interested in than Tesla’s spent since they were founded.
Musk is trying to solve the problem without LIDAR because he can't afford to use it. Necessity is the mother of his inventiveness. Advances in shrinking LIDAR arrays are continually happening, but LIDAR is still expensive as all get out. It's a deep pocket toy. Tesla is not a deep pocket company, not deep enough anyway.
Mozumder is wrong about self driving vehicles not happening. He is right partially right about infrastructure though. For self driving vehicles to become the norm, we don't need rail-on-road, we need wholesale improvements to our roads and bridges. A lot of our infrastructure is held together by duct tape and bubblegum. Self driving vehicles are going to need more consistent surfaces and markings to safely dominate the road.
Hmm, same was snails about iPod, iPhone, etc.
It’s Apple’s spin on wheels
Sure, i'm not saying Apple CANT expand into the car market in some way. I think it would be a great move to solidify revenues from alternative sources from pure tech.
But the moves / rumours we have seen so far over the years about "car" related stuff is just so jumbled and a mess and the market has seemingly moved faster than Apple seems capable of in this regards.
So the question I have still remains. What is their plan because as the years go on, everyone else in the car industry has seemingly moved faster and further on their own and don't seem all that interested in using someone else's tech. Apple would need to have something so dramatically different/ superior if they want to convince anyone else to use their self driving car tech (if that's what they're working on).
You hit it on the head when you stated you have no idea what Apple is doing. Here's just a hint though. Apple has close to a hundred self-driving cars driving on the roads every day and software putting millions of simulated miles, and has hired some of the top self-driving and automotive engineers in the industry, and has signed a partnership for a with Volkswagen to produce a van for their initial inter campus shuttle project, and has evidently hundreds of engineers working on Project Titan, FOR A REASON.
No reasonably informed person has ever suggested Apple intended to license software to car companies. They have never done that nor does it fit in any way with their business DNA or business model. Nor is make any sense with many companies developing such software, e.g., Google, Uber, etc., etc. What is much more likely is that Apple will design a complete vehicle and then having learned from Tesla's production nightmares, partner with a company, like Volkswagen, to produce it, provide maintenance and support, etc.
That company was here in Frisco TX, testing there Orange vans for several months, giving free rides between office parks. Most of the large boxes on the van were message boards to alert drivers of other cars. There was no issues with the van when they were here. There was a person in the front how ever just as safety per causation
DP
DP
Haha, you’ve never owned a Tesla clearly. I just drove from Oxford to Scotland without a hitch, I’d say around 70% of the driving was the car itself. Navigate on Autopilot rocks. Self driving cars here we come!
The smartest thing Apple did was to NOT buy Tesla. They have nothing of value except their brand name.
Everything you said seems will be a reality at some point, which is all under Mansfields leadership. Apple likely developing their _own_ software and possibly a vehicle design, but the partnership merges later with an auto manufacturer who will support the design manufacturing process. This whole ‘Apple Car’ project seems so far from fruition though, especially if you consider it’s been at least three years of continual rumors/side projects, ect. I question when Apple believes the time is/will be mature enough to bring whatever they have forward to the public.
Why do you think Tesla’s have such a high fatality rate?
[doublepost=1559831909][/doublepost]
So many people here about to die in their Teslas
[doublepost=1559831966][/doublepost]
Indeed. The government should ban Tesla’s like they do 737 Max!
Sure, that's what I mean by it being too expensive. He could afford to tinker with it (as he says they already use the tech at SpaceX), but why would he do that when he knows the market wouldn't pay the required premium for it to be profitable on their vehicles. He's a visionary and a designer but also a salesman and businessman.
I agree about needs for infrastructure improvements for the effectiveness of autonomy, especially in busy cities. However I also see what Tesla is doing to mitigate their dependence on surroundings and infrastructure (since it can be in constant flux, if nothing else because of construction and weather). For instance, instead of looking only at lane markings like they used to, they are learning to better identify driveable space and any objects therein, and I'm sure even pothole recognition is in the works. This would allow vehicles to operate much like humans do when the infrastructure is in bad repair or not clearly marked. It would also require that the car slow down and operate more cautiously to avoid incidents, but outside of busy cities, this approach would work just fine. Things like blind intersection turnouts and obstructed road signs will still need to be fixed, of course.
Did you forget /s? Please tell me you are joking, right? I hope...
An Apple self driving car will be technologically irrelevant every couple of years. That is Apple's DNA. They can't, for any reason, stick with anything for more than a few years. For a less then $1000 phone that's ok. For an $80,000 car not so much.
Driving entails engaging on a not-so-perfect world (and that is not going to change soon enough):
- road infrastructure -- or absence thereof.
- uncertain rules-of-the-road, when confronted with non-autonomy: trucks, cars, bikes, cyclists, jay-walkers...
- sudden weather change
If they do have a comparatively high fatality rate, which is debatable but admittedly hard to prove either way due to sample size differences, it's probably because many of the people who drive them are more prone to be dumb risk-takers. They either put too much trust in autopilot, or they drive more recklessly because the unusually high torque and low center of gravity allows them to.
There's no way you can claim that. There simply hasn't been enough data generated to form an opinion,.
This is why Musk made himself look like a complete and utter fool when that first fatality happened. There was 1 fatality and Teslas had driven 130 million miles on Autopilot to that point. Therefore Musk concluded Tesla had a fatality rate of 1 in 130 million miles on Autopilot, which was better than the average in the US of 1 fatality in 94 million miles.
Let's look at the numerous flaws with his statement.
- Sample Size: Anyone with even the most basic understanding of statistics knows the mistake Musk made. A single fatality is simply not enough data to make his claim. If there were 100 fatalities in 13 billion miles, then he would be justified in making his statement.
- Vehicle Age: The average age of motor vehicles in the US on the road is somewhere around 11-12 years. At the time Musk made his claim Teslas had only been on sale for 6 years, with the average age being even less than that. You can't compare the numbers of newer cars to that of older cars. Of course newer vehicles will have a lower fatality rate due to several factors, the most obvious being maintenance and state of repair.
- Safety Features: Musk likes to brag the Model S has a perfect 5 star safety rating (not quite true, but irrelevant for this discussion). The US fatality rate is for ALL vehicles from 5 star rated ones down to those with poor safety records. Vehicle safety (like airbags and crash worthiness) have a direct impact on fatalities, yet Musk is trying to attribute the safety of the Model S solely to Autopilot.
- Driving Conditions: Tesla Autopilot only works in good driving conditions. People don't use Autopilot in bad weather (rain/snow) or at night. The types of driving conditions where accidents are far more likely. So Musk is comparing the fatality rate of Teslas to that of vehicles that are driven in all types of situations where Teslas aren't driven.
- Vehicle Type: The Model S is a luxury class vehicle. The stats for road fatalities in the US he quoted are for all vehicles combined. If you separate out vehicles in the same class as a Model S (BMW 5 Series, Mercedes E Class and other mid-range luxury cars) you find the accident rate is actually 1 fatality in 320 million miles. Substantially better than Musks claim of 1 in 130 million miles for Tesla.
Bottom line, Musk should have never made his ridiculous claim in the first place, and we still don't have enough data to make an accurate prediction of the safety of Autopilot vs regular drivers (at least not until we can track miles driven in regular vehicles under the exact same limited conditions Autopilot works).
My question is why not release a self driving car with LIDAR? Okay if it costs $200,000. At least there is a starting point and someone will buy it. I know a lot of rich people who would drop $200,000 just for the status of being able to afford one.
NOT GOING TO HAPPEN. The American car companies are in cooperation with local governments to make the roads as crappy as possible so that the only way normal people can drive on them is via one of their SUVs or pick-up trucks.
We don't need infrastructure. A human can drive a vehicle with nothing more than a pair of eyes. I don't need high-res GPS maps stored in my brain, complex radar or other sensors feeding me information. I can drive at night, in rain or even on snow covered roads and my brain can figure out where I should be driving. I can navigate construction zones, traffic issues, accidents or other "surprises" with ease.
It's foolish to spend money on infrastructure (for self driving, not in general) when the technology/equipment you just installed on your roads to help with self driving will be obsolete in a few years by cars that no longer require those "aids" in order to function properly.
The other drawback to lidar, along with the cost, and why we haven't seen it on any cars for sale, is the size of the equipment. You can't fit it easily on the fringes of the car like you can with other sensors. The tech will shrink, but hasn't gotten to that point yet. I'd bet most people who plop down 200 large on a car don't want it to look like a science projectMy question is why not release a self driving car with LIDAR? Okay if it costs $200,000. At least there is a starting point and someone will buy it. I know a lot of rich people who would drop $200,000 just for the status of being able to afford one.
NOT GOING TO HAPPEN. The American car companies are in cooperation with local governments to make the roads as crappy as possible so that the only way normal people can drive on them is via one of their SUVs or pick-up trucks.
Wow, that statement is as bold as it is short-sided!!!
I’m grateful that we won’t have to wait too awful long before you’re handily proven incorrect.
But if they cost one life that wouldn't have been lost if a human were driving, the public will reject them. We need only look at the airline industry to see how high a standard the public holds automated vehicles. Right now Teslas are known for obliviously driving full speed into the side of semis. I doubt the public would accept such a vehicle if it were advertised as self-driving but as it is now, we can blame the driver for not monitoring the "driver assist" system.