Science

Elon Musk claims Tesla has created 'the best chip in the world' for autonomous vehicles


Elon Musk has given a rare glimpse into the underpinnings of his electric car company’s futuristic autonomous vehicle technology. 

At Tesla’s first-ever Autonomy Day with investors, the firm revealed it has developed what it says is the ‘best chip in the world’ that will allow its cars to achieve full self-driving capabilities without the need for human intervention. 

The new chip has allowed Tesla to make strides in bringing fully autonomous software to its fleet of vehicles, so much so that Musk predicts Tesla will have more than one million fully self-driving cars on the road by 2020. 

Scroll down for video 

Elon Musk has given a rare glimpse into the underpinnings of his electric car company's futuristic autonomous vehicle technology. The firm held its first-ever Autonomy Day

Elon Musk has given a rare glimpse into the underpinnings of his electric car company’s futuristic autonomous vehicle technology. The firm held its first-ever Autonomy Day 

‘It seems improbable. How could it be that Tesla, who has never designed a chip before, would design the best chip in the world?’ Musk said on stage at the event, which was hosted at Tesla’s Palo Alto, California headquarters. 

‘But that is objectively what has occurred. All Teslas being produced right now have this chip. 

‘All Tesla cars right now have everything necessary for full self-driving. All you have to do is update the software,’ he added.  

Peter Bannon, Tesla’s vice president of autopilot engineering, said the computer is low power and low cost, as well as ‘straightforward and simple.’

The chip has six billion transistors and can process 2.5 billion pixels per second from onboard cameras, Bannon said. 

Tesla claims it has developed the 'best chip in the world' that will allow its cars to achieve full self-driving capabilities without the need for human intervention

Tesla claims it has developed the ‘best chip in the world’ that will allow its cars to achieve full self-driving capabilities without the need for human intervention

The firm said the computer is low cost and low power, as well as 'straightforward and simple.' It is responsible for processing and receiving data from the car's eight, 360-degree cameras

The firm said the computer is low cost and low power, as well as ‘straightforward and simple.’ It is responsible for processing and receiving data from the car’s eight, 360-degree cameras 

The self-driving computer is responsible for processing and receiving data from the car’s eight, 360-degree cameras. 

There are two chips integrated in a single computer inside the car, both of which collect information from the world around you, using a combination of mapping, radar and ultrasonic sensors, Bannon explained. 

The two machines exchange their analysis of the data, they ‘agree then act’ and inform the car’s decision-making process. 

Musk said there’s ‘a tremendous amount of redundancy and overlap’ to make sure the computer operates as safely as possible. 

‘Any part of this could fail and the car keeps driving,’ he explained. ‘The probability of this computer failing is substantially lower than a human losing consciousness.’ 

The computer has six billion transistors and can process 2.5 billion pixels per second from onboard cameras. There's a 'tremendous amount of redundancy' to improve safety

The computer has six billion transistors and can process 2.5 billion pixels per second from onboard cameras. There’s a ‘tremendous amount of redundancy’ to improve safety

There are two chips integrated in a single computer inside the car, both of which collect data from the world around you, using a combination of mapping, radar and ultrasonic sensors

There are two chips integrated in a single computer inside the car, both of which collect data from the world around you, using a combination of mapping, radar and ultrasonic sensors

The company is now working on developing its next-generation chip, which Musk says is 'at least' three times better than the current version. He expects it to be ready in two years

The company is now working on developing its next-generation chip, which Musk says is ‘at least’ three times better than the current version. He expects it to be ready in two years

Tesla’s new chip represents a significant performance change from the previous model, Musk said. 

It can now run neural networks that are seven times larger and ‘more sophisticated.’ 

Andrej Karpathy, Tesla’s director of AI and Autopilot vision, explained just how challenging it is to train self-driving vehicles to handle all the scenarios encountered on the road. 

The company has been working for at least five years to make its neural networks more intelligent. 

This involves humans annotating and labeling data to help Tesla computers recognize common things like bicycles and tractor trailers, as well as common behaviors, such as when a car is changing lanes. 

‘As you add more data, neural networks start to work better for free,’ Karpathy said. 

Critics have pointed out that Tesla’s autonomous software doesn’t use LIDAR – a technology that’s become commonplace in conversations about self-driving cars. 

LIDAR relies on laser sensors to assist with the car mapping its surroundings. 

‘LIDAR is a fool’s errand,’ Musk said. ‘Anyone relying on LIDAR is doomed. 

‘…It’s expensive, and unnecessary and as Andre was saying, once you solve vision, it’s worthless to have expensive hardware on the car.’ 

The company is now working on developing its next-generation chip, which Musk says is ‘at least’ three times better than the current version.  

Musk added that he expects the new chip to be ready in about two years. 

HOW DO SELF-DRIVING CARS ‘SEE’?

Self-driving cars often use a combination of normal two-dimensional cameras and depth-sensing ‘LiDAR’ units to recognise the world around them.

However, others make use of visible light cameras that capture imagery of the roads and streets. 

They are trained with a wealth of information and vast databases of hundreds of thousands of clips which are processed using artificial intelligence to accurately identify people, signs and hazards.   

In LiDAR (light detection and ranging) scanning – which is used by Waymo – one or more lasers send out short pulses, which bounce back when they hit an obstacle.

These sensors constantly scan the surrounding areas looking for information, acting as the ‘eyes’ of the car.

While the units supply depth information, their low resolution makes it hard to detect small, faraway objects without help from a normal camera linked to it in real time.

In November last year Apple revealed details of its driverless car system that uses lasers to detect pedestrians and cyclists from a distance.

The Apple researchers said they were able to get ‘highly encouraging results’ in spotting pedestrians and cyclists with just LiDAR data.

They also wrote they were able to beat other approaches for detecting three-dimensional objects that use only LiDAR.

Other self-driving cars generally rely on a combination of cameras, sensors and lasers. 

An example is Volvo’s self driving cars that rely on around 28 cameras, sensors and lasers.

A network of computers process information, which together with GPS, generates a real-time map of moving and stationary objects in the environment.

Twelve ultrasonic sensors around the car are used to identify objects close to the vehicle and support autonomous drive at low speeds.

A wave radar and camera placed on the windscreen reads traffic signs and the road’s curvature and can detect objects on the road such as other road users.

Four radars behind the front and rear bumpers also locate objects.

Two long-range radars on the bumper are used to detect fast-moving vehicles approaching from far behind, which is useful on motorways.

Four cameras – two on the wing mirrors, one on the grille and one on the rear bumper – monitor objects in close proximity to the vehicle and lane markings. 



READ SOURCE

Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.