The video of the 29-ton truck moving in tight spaces full of Chinese vases and the video of Tesla hitting a child’s puppet (which Musk claims to be false) suddenly rekindle the controversy over the “eyes” of self-driving cars. The sensors that allow the machine to interpret what happens on the road and therefore to move the steering, accelerator and brakes.
Tesla does not stop and hits a child’s dummy: the test leads to discussion
Yes, because while the whole automotive world – including NASA – uses Lidar sensors, Elon Musk disputes them and does not use them on Tesla (which only go ahead with normal cameras). A tough dispute: Musk even defined Lidar technology as “stupid”. Adding that “Anyone who relies on Lidar is doomed.”
To understand, the truck that avoids the Chinese vases by a whisker used the sensors hated by Musk, the Model 3 that hit the fake baby did not. Hence the controversy.
But how does Lidar work? Quite complexly: when the light emitted by the laser hits the object, the part of it that comes back is collected by a sensor and computed by a processor to generate a three-dimensional reconstruction of what the car is facing. The digital reconstruction of the surrounding environment is perfect.
But then why does Musk hate him? For two reasons. The first is that it is very expensive. The second, on the other hand, is almost philosophical. Musk himself explained that “mounting Lidars on the car means filling it with expensive appendages. But on a car every added accessory is a bad thing: it is ridiculous to fill the car with these devices.
How does a 29-ton truck move by itself in a maze of precious Chinese vases
Furthermore, there is another aspect, as we said “philosophical” in Elon Musk’s approach to this technology: according to him, we start from a wrong premise: “We drive by processing visual images – he explained – and the self-driving car must do the same. In other words, interpreting the images ”.
Privacy on wheels
Cybersecurity on the move, or travel safely with the cars of the future
by Luca Indemini
In fact, Lidar advocates say this technology can do things that cameras absolutely cannot do. And – from a technical point of view – they are essentially right because normal cameras interpret depths, the Lidar instead measures all distances with millimeter precision.
In short, the debate is open. And he has come to blow the controversy on the fire the sighting – complete with a Twitter train – of a prototype of a Tesla Model Y equipped with Lidar systems in Florida.
Musk’s company was quick to state that it was just a test, but in reality, given the enormous technical progress of Lidar systems, the doubt remains. Yes, because in the meantime the Lidar system has reached the third generation and promises wonders: it has a resolution about 50 times higher than the second generation and in fact it sees everything, even when the object is very far away, invisible to the human eye.
It can identify objects over 150 meters away, contrary to what an eye, a radar or a camera can do, like a small object with very low reflectivity, such as a tire. Not only that: it is able to recreate a 3D image of the space surrounding the vehicle using a point cloud, with a resolution still unsurpassed for an automotive system. It can then map the terrain topology and detect road signs. All in all climatic conditions (from -40 ° C to + 85 ° C).
The world‘s first manufacturer of Lidar is laValeo (turnover of € 17.3 billion and invests 8.7% of turnover in Research & Development, with 184 production plants, 21 research centers, 43 development centers, 16 distribution platforms and 103,300 employees), already number one in the world in ADAS (Advanced Driver Assistance Systems), the driver assistance systems.
Valeo was the first, and today the only one, to produce an automotive Lidar scanner on an industrial scale: over 170,000 units were produced and the technology is protected by more than 500 patents. Among these is also the one for high-performance embedded software based on artificial intelligence algorithms, which allow it to change the trajectory of the vehicle, anticipating the areas without obstacles on the road. He can self-diagnose and trigger his cleaning system when his field of vision is obstructed. Here, by studying how the third generation of the Lidar works, we can probably find an answer to the mysterious images of the Tesla Model 3 with these systems on the roof.