Waymo and the simulation for autonomous cars

It has always been known that Waymo uses AI tools to simulate the behavior of its autonomous cars and thus improve their operability, but recently it has released the name of one of its simulation programs called Simulation City, and it represents a city through which the autonomous car circulates and in which takes place all kinds of possible situations.

Waymo has more than 600 autonomous cars operating in various North American cities and performs on the order of 20 million miles each year, but its simulation programs run on the order of 15,000 million miles. Almost 1000 times more, and that means that there is a lot of analysis work that is not seen, but that ends up improving the programs that control the operation of the autonomous cars in tests.

Lately the public’s enthusiasm for these types of companies has waned, when they have realized that things are not as easy as they were promised, and some companies like UBER have left the market. But I think they have a long-term future and perhaps in specific areas of the total automotive market.

With the Simulation City program, engineers can simulate almost anything, from the effect of the sun’s rays at dawn on vision cameras and how they can perceive traffic signs, or the “crazy” behavior of a driver on the road.

And all of that adds to the knowledge that Waymo has long had.

That is why their cars are the ones that have had the fewest accidents despite being the ones that have made the most kilometers.

It is very difficult to predict the future, but I see Waymo with its “driving” technology collaborating with car companies to jointly offer autonomous vehicles within a few years.

When and how is yet to be seen.