Solid state Lidar progressing

[lolder]

The percentage confidence is in the upper left corner of the box. They're hard to read.

 

[electronblue]

Reality is it's easy to say things (that were never done before) can not be done and you will more often be right than wrong. Those few who say things can be done are ones who made the world what it is today. We should be very careful in saying who is "right" and who is not that way. The right way to do analysis is with physics but not anecdotal evidences which is what most people, especially naysayers, do. Kind of like what Elon said the right way to do engineering design by the first priciple and not by analogy.

I think there are two points that may be relevant here:

1. Elon Musk may have made an uncharacteristic error in judgement with FSD, especially by starting sales and making the promises that he did. It does not really take a negative person to question that move.

2. @Bladerskb probably works in the industry that actually makes autonomous driving stuff so his/her opinion probably has more value from that perspective not just because it is easy to be negative. It also explains why he/she had the foresight many amateurs lacked.

This is one alternative perspective to who and why got this right.

 

[CarlK]

2. @Bladerskb probably works in the industry that actually makes autonomous driving stuff so his/her opinion probably has more value from that perspective not just because it is easy to be negative. It also explains why he/she had the foresight many amateurs lacked.

This is one alternative perspective to who and why got this right.

Note that I mentioned the first priciple of engineering design? People who work in the industry could be more prone to follow analogy rather than the first priciple. Just take a deep look at everything Elon did versus everything the industry is doing you'll see clearly why.

1. Elon Musk may have made an uncharacteristic error in judgement with FSD, especially by starting sales and making the promises that he did. It does not really take a negative person to question that move.

Using this as a good example of Elon's first priciple he rejected things others are doing, most noteworthy LIDAR based system everyone is copying Waymo. That would be a obvious way to go (by analogy) but I'm sure he did not see a clear path to get there. People today probably could not see the clear path either other than let's start it and see if we can have this or that, such as practical LIDAR you can put in every passenger car, in the future. They are not going to get anywhere if the path choosen is wrong.

I could eleborate on why Elon's approach conforms to the first priciple and why it has the best chance of success. It will be too long a post and I've saiy many times why that approach is superior. I might find time in the future to make my arugument again if anyone is interested to listen.

Regardless my opinion of Mobileye is dishonest in presenting its technology is unrelated to this.

 

[renim]

Mid LiDAR specs- Livox

This pretty much settles it, going forward every autonomous car will have at least 1 LIDAR. There are so many uses for this, not just in cars and transport, this is big, similar to cost effective GPS.

 

[electronblue]

Note that I mentioned the first priciple of engineering design? People who work in the industry could be more prone to follow analogy rather than the first priciple. Just take a deep look at everything Elon did versus everything the industry is doing you'll see clearly why.

Using this as a good example of Elon's first priciple he rejected things others are doing, most noteworthy LIDAR based system everyone is copying Waymo. That would be a obvious way to go (by analogy) but I'm sure he did not see a clear path to get there. People today probably could not see the clear path either other than let's start it and see if we can have this or that, such as practical LIDAR you can put in every passenger car, in the future. They are not going to get anywhere if the path choosen is wrong.

I could eleborate on why Elon's approach conforms to the first priciple and why it has the best chance of success. It will be too long a post and I've saiy many times why that approach is superior. I might find time in the future to make my arugument again if anyone is interested to listen.

The problem with many things, first principles included, is that in the end they are fodder for philosophical debate... and especially their application comes down to opinions. I also think that first principles probably is at its strongest when questioning established industries where iterative processes usually reign supreme but autonomous driving hardly is an established industry.

What does it take to drive a car? Sensors that sense, decision-making that chooses and mechanics that drive. These same questions can well lead you down to emphasizing different parts differently, but that does not automatically in practice lead to a best result in some particular direction.

For example saying that eyes and a large brain work because they work for humans would be analogous, not first principles. First principles actually suggest you question that analogy. Maybe super-human sensors (many times over) and a miniscule brain would do it better.

Regardless my opinion of Mobileye is dishonest in presenting its technology is unrelated to this.

I myself find MobilEye much more honest than Tesla in presenting its autonomous-related technology.

 

[caligula666]

Using this as a good example of Elon's first priciple he rejected things others are doing, most noteworthy LIDAR based system everyone is copying Waymo. That would be a obvious way to go (by analogy)

Actually, going with just camera and computer is the closest you can do from an analogy to a human. So not sure I agree with your classification here.

Plus some tend to forget Tesla has radars. We are already past the pure camera situation, both in sensors, but also in the sense of having to solve possibly conflicting signals. So LIDARs wouldn’t bring that baggage, it is already there.

 

[CarlK]

Actually, going with just camera and computer is the closest you can do from an analogy to a human. So not sure I agree with your classification here.

Plus some tend to forget Tesla has radars. We are already past the pure camera situation, both in sensors, but also in the sense of having to solve possibly conflicting signals. So LIDARs wouldn’t bring that baggage, it is already there.

If by analogy he would be using LIDAR like everyone else does because the only system in existence was using it. Eyes and brain is just the result of nature's way of applying the first principle. The coincidence is not unexpected. As for radar I believe Elon mentioned it only plays a minor role and even that role may one day deem unnecessary.

 

[electronblue]

As for radar I believe Elon mentioned it only plays a minor role and even that role may one day deem unnecessary.

This I agree with. I believe mostly Tesla’s radar and ultrasonics in Autopilot 2-3 are leftovers from Autopilot 1 that have enabled TACC, Summon and Autopark during the time Tesla Vision was/is still maturing and Tesla couldn’t implement those features through vision. After all, EyeQ3 was originally supposed to be there on Autopilot 2 board too.

It remains to be seen what if any ”redundancy” they will provide over time or if Tesla truly goes vision only. They might. There is especially one scenario where I question their ability to go vision only though and that is parking and tight maneuvers forward. Do the camera blindspots cause a need to rely on sensor fusion in those scenarios? They might.

Of course it is possible Tesla changes this the next time they change the suite but so far this is what it looks like to me. Without bumper heating the forward radar is not that reliable anyway in many regions and it only sees a narrow cone forward, nothing to the sides for example which would be useful for cross-traffic.

 

[Bladerskb]

Regardless my opinion of Mobileye is dishonest in presenting its technology is unrelated to this.

Wait Mobileye is dishonest but the guy who sold his car and claimed multiple times on every occasion that it will be fully autonomous and require no human driver including being summoned from across the country by end of 2017/early 2018 isn't ? Knowing full well that his comments which generate hundreds of articles leading directly to tens of thousands of purchases for the very reason that it will be fully autonomous?

 

[CarlK]

Wait Mobileye is dishonest but the guy who sold his car and claimed multiple times on every occasion that it will be fully autonomous and require no human driver including being summoned from across the country by end of 2017/early 2018 isn't ? Knowing full well that his comments which generate hundreds of articles leading directly to tens of thousands of purchases for the very reason that it will be fully autonomous?

The spin Mobileye put out during the Tesla separation, now that we know the history better, was indeed dishonest. What they said in 2016 and later about their machinie learning commitment were totally contrary to each other. They're just making up stories as it goes. Elon has been consistent with what he said he plans to do. The only fault is his timeline is overly optimistic but that's the way he's always been, overly optimistic, and nothing dishonest about it.

 

[Bladerskb]

The spin Mobileye put out during the Tesla separation, now that we know the history better, was indeed dishonest. What they said in 2016 and later about their machinie learning commitment were totally contrary to each other. They're just making up stories as it goes. Elon has been consistent with what he said he plans to do. The only fault is his timeline is overly optimistic but that's the way he's always been, overly optimistic, and nothing dishonest about it.

saying the impossible is guaranteed to happen over and over again is not overly optimistic its fraud!
saying you will have AP1 parity in 3 months after AP2 launch when not one single line of code had been written and you knew it is fraud. Saying that EAP will be here in Dec 2016 again is fraud.

You think his engineers are stupid? That they are just throwing out these lofty timelines and that's where elon is getting it from? wake up!

There's a difference between overly optimistic and straight up lying. selling cars for 100k based on that is as dishonest as it can get.
Not only that but they have already been sued by the said fraud and lost the suit. And then we know about the 420 fraud aswell. So you are bashing mobileye for being dishonest yet praising Elon for his truthfulness? A guy who came out and claimed a random guy whom he doesn't know was a pedo and had child brides and that he had proof? This is your model of honesty?

Elon in your mind literally can do no wrong.

 

[Electroman]

And then we know about the 420 fraud aswell

Okay.. now you are showing your colors.. stick to AP related discussions and you will find many people who will take your seriously.

 

[Bladerskb]

Okay.. now you are showing your colors.. stick to AP related discussions and you will find many people who will take your seriously.

Stupid auto-correct, I meant to say.. I have no respect for the SEC
They have showed their true colors, Total Witch Hunt!