Artificial intelligence was at the heart of the last CES show in Las Vegas. The technological innovations of recent years have changed the way we drive by enhancing the vehicles with driving aid tools. But these tools are limited to the automation of tasks such as parking or braking assistance. Human-beings remain essential to coordinate these different tools.
Today, we are approaching a new technological milestone. The ambition is no longer to simply make driving safer or to make it more enjoyable. From now on, we need to focus on artificial intelligence to achieve the ultimate stage of transformation: the self-driving vehicle.
With A.I., vehicles will no longer be content on passively receiving orders from the drivers. They will become reactive, able to adapt their behaviour to the conditions of the road at the same level of human vigilance. The challenge is to find a system able to anticipate and analyse the situation to offer solutions without taking any risks.
The company INTEL considers the A.I. sector to be a strategic priority for their development.
- However, before benefitting from the advantages of A.I., engineers and industrialists have to deal with a number of major issues. To allow the computer to analyse each situation, you need to create an infrastructure that can handle very important data streams. An hour of driving in a self-driving vehicle generates 4 TB of data. This represents the equivalent of 3,000 people surfing the internet over 24 hours. One million self-driving cars represents as much data as the entire world population connected to the Internet.
- After analysing the situation, the computer must be able to make a decision. The computational capabilities necessary to achieve a viable and safe solution requires developing a computer capable of making a colossal volume of calculations – over 30 billion different operations. This capacity is equivalent to three human brains.
- I. in vehicles relies on the hyper connectivity of the car. Vehicles will have to communicate continuously with the environment, with other automobiles, the road infrastructure and geolocation systems. To ensure sufficient accuracy the onboard computer will have to be able to validate its position on the road to the nearest centimetre, whatever the climatic conditions. The development of A.I. therefore depends on the emergence of new communication systems such as the 5G.
A.I. is the equivalent of a computer with the capacity of three human brains.
We still have a way to go before we can have vehicles that are able to suggest a route according to our mood and to adapt its driving according to our state of fatigue. In addition, some market analysists are not convinced that this is what consumers want. There is always a kind of reluctance to accept advances in technology for fear that it will dominate us. Man wants more technology as long as it’s not smarter than he is!
Discover our file on self-driving vehicles Https://mylab.opteven.com/index.fhp/catégorie/dossiers-grand-angle/vehicules-autonomes/