[ad_1]
<div _ngcontent-c14 = "" innerhtml = "
Autonomous self-driving car without driver, should it have a LIDAR or not? That is the question.
Getty
Readers asked me to supplement my remarks about Tesla & nbsp;Investor Independence Day& nbsp; in which I had indicated Elon Musk resolutely and antagonist threw the glove to strengthen his position by questioning the use of LIDAR (a type of sensor using light and radar). In fact, he threw the glove off & nbsp; directly to the forefront of the autonomous car sector and those who do & nbsp; the devices.
In short, Musk has become unequivocally the de facto head of anti-LIDAR camp.
Do not forget that almost every other driverless car manufacturer incorporates the use of LIDAR in their vision systems. In addition, technology companies that already manufacture sensors or develop LIDAR units have recently flourished, with more than 80 innovative companies (according to SAE). Automotive Engineering Magazine indication), and drew without shame Silver VC / PE easily.
You have a voracious trumpet player who runs in the hills that LIDAR is verboten, adopting a manifest and highly visible anti-LIDAR posture, then pretty much everyone embracing LIDAR for autonomous cars. And even if sometimes a contrarian may be right, they are often as likely to be wrong, maybe more.
It's now become an undoubtedly considerable bet on the part of Musk and a gigantic gamble that could ultimately undermine Tesla's chances of survival, which I will explain here.
How did we come here
Before his extremely vitriolic anti-LIDAR proclamation, Musk had generally been a little more discreet about his position (silent is a relative term).
Looking back in the past, early in 2018, we may have considered dogs a year or so ago, and we realized that he had argued that LIDAR was excessively expensive and bulky and that he regarded it as a kind of crutch, even though he had conceded"Maybe I'm wrong and I'm going to look silly. But I am quite certain that I am not. "
On April 22, 2019 (Tesla Investor Independence Day), the world seemed to be a completely different place. As far as Elon's opinion on LIDAR is concerned, let's say, let's say, less reserved and more categorical, he proclaimed: "LIDAR is a crazy race. Anyone who uses LIDAR is doomed. Condemned!
Slightly, one could say that he chose to double his bet.
A little earlier, he had only one cover, offering the possibility that the LIDAR is useful. He has now become the anti-LIDAR proctor, using his speaker to declare LIDAR null and void. Worse still, his comments claim that others who use or invest in LIDAR for autonomous cars are reckless, presumably losing money, time, energy and their minds on a doomed technology (according to him ).
Musk often takes air in a room, but this one was particularly hectic, leaving others in the auto industry gasping for breath, although, given the often used hyperbole resulting from Tesla and Musk he was not a surprise for those who are already accustomed to its sometimes scandalous (bold?) decrees.
What is lidar?
Clarify, LIDAR was this bulbous type hood that you no doubt saw sitting at the top of autonomous cars.
It emits light beams, the light beams bounce on nearby objects, the return beams are then collected by the LIDAR unit. This is similar to a radar and allows the measurement of distances to objects, in the range of the LIDAR device used. The artificial intelligence system of the autonomous car then interprets the collected data to try to determine which objects are nearby, as well as their shapes, their amount of movement over time, their direction or direction, and so on.
When most autonomous car manufacturers started with designing driverless cars, they tended to include LIDAR units in their general AI system for cars. Musk is right about the idea that LIDAR units are expensive and bulky, but it's a bit like living in the past, because advances in technology have dramatically reduced costs and sizes are also much smaller. In fact, many driverless cars are now equipped with multiple LIDAR units and you can hardly notice them.
Nowadays, it is much more difficult to support the anti-LIDAR argument by claiming that LIDAR units are cumbersome and costly. It's one of those arguments that really does not hold water anymore, although some people still cling to it anyway. Leaving aside this part of the argument as torn apart and less relevant.
Instead, consider an ongoing technical argument in the halls of automakers and independent researchers about the advantages of LIDAR over cameras and the anti-LIDAR camp that claims that cameras are sufficient and that LIDAR does not is not necessary.
In essence, some say that since humans only use their eyes when they drive and can not emit light beams, it may mean that only the cameras are sufficient and that the LIDAR is useless ( the classic replica is that humans only have legs, so presumably cars should have legs rather than wheels, pointing out that the mere fact of imitating human features is not really a valid argument in itself).
What immediately compromises the fundamental point of the cameras is that almost everyone seems to agree to this day that the use of radar is an essential partner in the use of cameras. Even Musk allowed this to be the case of Tesla and indeed Tesla cars are equipped with a radar. If this is the case, then the question of whether to use a conventional radar rather than LIDAR becomes a debate, since the door has already been opened to agree that this type of radar is justified, rather than worrying. LIDAR dialogue against cameras.
To plunge deeper into the acrimonious debate, the usual response from the all-camera camp is that conventional radar is just a temporary bridge, and once the cameras are good enough, the radar will no longer be visible. necessary or will be considered a mere convenience. availability. They often throw in mind the notion that the radar costs less than the LIDAR and is less cumbersome, but as I mentioned earlier, these flippant comments do not stand the test of time.
Why this anti-lidar is a heavy bet
I say that these recurring arguments lack quite fatal aspects that weigh even more on the future of autonomous cars and especially driverless cars in the anti-LIDAR encampment.
When an autonomous car enters an accident, which will mark my words will arrive, you can rest assured that proceedings will be instituted. I've already described that one of the key elements of such trials will be what the independent car manufacturer did in the design, construction and commissioning their autonomous car, and in particular how did they seek to ensure safety and reliability.
For Tesla, if they go to trial, they will have to defend in court their decision not to use LIDAR. As you can imagine, Tesla will be on rather unstable ground if it is shown that all other autonomous cars use LIDAR. The burden of explaining and justifying the absence of LIDAR on Tesla will be extremely heavy.
And more, I reported several times Musk's claim that the cost of LIDAR is "expensive" will put Tesla in even more troubled waters, putting a cost-related issue at the forefront.
Assuming that people have died in the car accident, the question of the cost of a LIDAR unit at the time and its integration into a driverless car will be clearly posed, relative to the cost lost human lives. Traditional builders are well aware that juries and judges have a bleak view of sensitive matters that can be attributed to a less secure or less reliable vehicle (remember the well-known Pinto case and rumors about the cost of safety versus cost in human lives).
Tesla's position would probably be that the addition of the LIDAR would not have materially avoided the car accident and the loss of life, but this will be difficult to demonstrate because, in theory, any use of LIDAR will gradually improve the chances in the following hypothesis: It is used wisely. Tesla therefore avoids being summoned by the absence of LIDAR.
Nor can they claim that they did not know LIDAR or did not know it one way or the other, which is obviously not the case, including the fact that their anti-LIDAR rhetoric that they had themselves presented was their avowed admission that they had heard about LIDAR and had a deliberate decision to exclude it voluntarily.
Conclusion
Even if Musk were to change his mind and opt for LIDAR camp, Tesla repeated several times that his cars had already Necessary material for autonomous abilities. This implies that Tesla would be ready to equip Tesla with LIDAR, after the fact, by rearranging them all, a very expensive and unimaginable event logistically, if they had an epiphany and a radical change of heart about LIDAR.
In short, Tesla is now painted in a narrow corner.
They can not easily get out of the LIDAR mode to enter the LIDAR adopting group, and when the time comes to defend themselves in court, it will be easier to explain to the jurors and judges that Tesla has omitted something that everyone uses and for which he could have helped to avoid the accident and the deaths, compared to Tesla who tried to argue that they had done very well without the LIDAR (more, bothered to explain why they had saved on costs, which implied some sort of – the calculation of life was involved).
If this happens and goes against Tesla, or even if it only raises sensitive questions about the safety of Tesla cars in the absence of LIDAR, the situation could be uneasy. public opinion on driverless cars that lack LIDAR and even spark regulators in action.
Anyway, Musk and Tesla bet a pretty interesting bet and the roll of the dice is already underway.
">
Autonomous self-driving car without driver, should it have a LIDAR or not? That is the question.
Getty
Readers asked me to supplement my remarks about Tesla's Investor Independence Day, in which I had told Elon Musk resolutely and antagonistly to launch the challenge to strengthen his position. by questioning the use of LIDAR (a type of sensor using light and radar). In fact, he launched the challenge to face the rest of the autonomous auto industry and those who manufacture the devices.
In short, Musk has become unequivocally the de facto leader of the anti-LIDAR camp.
Keep in mind that almost every other driverless car manufacturer incorporates the use of LIDAR in their vision systems. In addition, technology companies that already manufacture sensors or develop LIDAR units have recently flourished, bringing together over 80 innovative companies (according to SAE's Automotive Engineering magazine) and have easily attracted venture capital funds. capital investment.
You have a voracious trumpet player who runs in the hills that LIDAR is verboten, adopting a manifest and highly visible anti-LIDAR posture, then pretty much everyone embracing LIDAR for autonomous cars. And even if sometimes a contrarian may be right, they often tend to err, perhaps even more.
It's now become an undoubtedly considerable bet on the part of Musk and a gigantic gamble that could ultimately undermine Tesla's chances of survival, which I will explain here.
How did we come here
Before his extremely vitriolic anti-LIDAR proclamation, Musk had generally been a little more discreet about his position (silent is a relative term).
Looking back in the past, early in 2018, we may have considered dogs a year or so ago, and we realized that he had argued that LIDAR was excessively expensive and bulky and that he regarded it as a sort of crutch, although he also conceded: "Maybe I'm wrong and I'm going to look like one." imbecile. But I am quite certain that I am not. "
On April 22, 2019 (Tesla Investor Independence Day), the world seemed to be a completely different place. With regard to Elon's opinion on LIDAR, he has become less categorical and more categorical. Here's what he proclaimed: "LIDAR is a crazy race. Anyone who uses LIDAR is doomed. Condemned!
Slightly, one could say that he chose to double his bet.
A little earlier, he had only one cover, offering the possibility that the LIDAR is useful. He has now become the anti-LIDAR proctor, using his speaker to declare LIDAR null and void. Worse still, his comments claim that others who use or invest in LIDAR for autonomous cars are reckless, presumably losing money, time, energy and their minds on a doomed technology (according to him ).
Musk often takes air in a room, but this one was particularly hectic, leaving others in the auto industry breathless, although, given the often used hyperbole resulting Tesla and Musk, it was not a surprise to those who already used it. to its sometimes scandalous decrees (audacious?).
What is lidar?
To clarify, LIDAR has been this bulb-like plug that you have no doubt seen sitting on top of autonomous cars.
It emits light beams, the light beams bounce on nearby objects, the return beams are then collected by the LIDAR unit. This is similar to a radar and allows the measurement of distances to objects, in the range of the LIDAR device used. The artificial intelligence system of the autonomous car then interprets the collected data to try to determine which objects are nearby, as well as their shapes, their amount of movement over time, their direction or direction, and so on.
When most autonomous car manufacturers began designing driverless cars, they tended to include LIDAR units in their general AI system for cars. Musk is right about the idea that LIDAR units are expensive and bulky, but it's a bit like living in the past, because advances in technology have dramatically reduced costs and sizes are also much smaller. In fact, many driverless cars are now equipped with multiple LIDAR units and you can hardly notice them.
Nowadays, it is much more difficult to support the anti-LIDAR argument by claiming that LIDAR units are cumbersome and costly. It's one of those arguments that really does not hold water anymore, although some people still cling to it anyway. Leaving aside this part of the argument as torn apart and less relevant.
Instead, consider an ongoing technical argument in the halls of automakers and independent researchers about the advantages of LIDAR over cameras and the anti-LIDAR camp that claims that cameras are sufficient and that LIDAR does not is not necessary.
In essence, some say that since humans only use their eyes when they drive and can not emit light beams, it may mean that only the cameras are sufficient and that the LIDAR is useless ( the classic replica is that humans only have legs, so presumably cars should have legs rather than wheels, pointing out that the mere fact of imitating human features is not really a valid argument in itself).
What immediately compromises the fundamental point of the cameras is that almost everyone seems to agree to this day that the use of radar is an essential partner in the use of cameras. Even Musk admitted that this was the case for Tesla and, in fact, Tesla cars are equipped with a radar. If this is the case, then the question of whether to use a conventional radar rather than LIDAR becomes a debate, since the door has already been opened to agree that this type of radar is justified, rather than worrying. LIDAR dialogue against cameras.
To plunge deeper into the acrimonious debate, the usual response from the all-camera camp is that conventional radar is just a temporary bridge, and once the cameras are good enough, the radar will no longer be visible. necessary or will be considered a mere convenience. availability. They often throw in mind the notion that the radar costs less than the LIDAR and is less cumbersome, but as I mentioned earlier, these flippant comments do not stand the test of time.
Why this anti-lidar is a heavy bet
I say that these recurring arguments lack quite fatal aspects that weigh even more on the future of autonomous cars and especially driverless cars in the anti-LIDAR encampment.
When an autonomous car falls into an accident, which indicates that my words will happen, you can rest assured that lawsuits will be filed. I've already described that one of the key elements of these lawsuits would be what the auto maker did as part of the design, construction and commissioning of his car autonomous, and in particular how have they sought to ensure safety and security. reliability.
For Tesla, if they go to trial, they will have to defend in court their decision not to use LIDAR. As you can imagine, Tesla will be on rather unstable ground if it is shown that all other autonomous cars use LIDAR. The burden of explaining and justifying the absence of LIDAR on Tesla will be extremely heavy.
And even more, I have repeatedly pointed out that Musk's claim that the cost of LIDAR is "expensive" would put Tesla in even darker waters, putting a cost-related issue first.
Assuming that people have died in the car accident, the question of the cost of a LIDAR unit at the time and its integration into a driverless car will be clearly posed, relative to the cost lost human lives. Traditional builders are well aware that juries and judges have a bleak view of sensitive matters that can be attributed to a less secure or less reliable vehicle (remember the well-known Pinto case and rumors about the cost of safety versus cost in human lives).
Tesla's position would probably be that the addition of the LIDAR would not have materially avoided the car accident and the loss of life, but this will be difficult to demonstrate because, in theory, any use of LIDAR will gradually improve the chances in the following hypothesis: It is used wisely. Tesla therefore avoids being summoned by the absence of LIDAR.
Nor can they claim that they did not know LIDAR or did not know it one way or the other, which is obviously not the case, including the fact that their anti-LIDAR rhetoric that they had themselves presented was their avowed admission that they had heard about LIDAR and had a deliberate decision to exclude it voluntarily.
Conclusion
Even if Musk were to change his mind and opt for LIDAR camp, Tesla repeated several times that his cars already had the necessary equipment for autonomous capabilities. This implies that Tesla would be ready to equip Tesla with LIDAR, after the fact, by rearranging them all, a very expensive and unimaginable event logistically, if they had an epiphany and a radical change of heart about LIDAR.
In short, Tesla is now painted in a narrow corner.
They can not easily get out of the LIDAR mode to enter the LIDAR adopting group, and when the time comes to defend themselves in court, it will be easier to explain to the jurors and judges that Tesla has omitted something that everyone uses and for which he could have helped to avoid the accident and the deaths, compared to Tesla who tried to argue that they had done very well without the LIDAR (more, bothered to explain why they had saved on costs, which implied some sort of – the calculation of life was involved).
If this kind of case occurs and goes against Tesla, or even if it simply raises tricky questions about the safety of Tesla cars in the absence of LIDAR, it could awaken public opinion about driverless cars lack of LIDAR and even spark action.
Anyway, Musk and Tesla bet a pretty interesting bet and the roll of the dice is already underway.