There’s no question as to whether or not self-driving cars would become mainstream. According to vehicle systems expert Antonio Lopez, the question isn’t if, but when.
Even with all of Lopez’s confidence, there are still many challenges at hand before driverless cars become the new thing. One challenge is teaching cars how to identify situations to learn how and when to react.
For that reason, The Autonomous University of Barcelona’s Computer Vision Center made Synthia, where Lopez leads the advanced driver assistance systems.
Synthia is a simulator that uses different neural networks and extensive learning to enhance how vehicle artificial intelligence systems control what’s going on around them, especially in unpleasant weather. It all began eight years ago before Google displayed their self-driving car.
According to Lopez, the group uses commercial video games to detect pedestrians. He says that with the sensors they use, they can now see what is in each pixel. The team can also determine how close the objects are from the camera which they consider vital information for the system’s vision.
Artificial intelligence vehicles are being taught with a wide variety of real-life pictures so that they can recognize different elements. They can also decipher between obstacles such as the road from the sidewalk on days with extreme weather. The software can understand images from the car’s camera allowing it to react accordingly.
Lopez also states that his team ran tests using models to recognize pedestrians with an autonomous car made by the Autonomous University of Barcelona.
Data from the simulator can help the scientific community in fantastic ways, but there is still a long way to go.
Lopez believes that in 2050, approximately 75 percent of the population will reside in metropolitan areas. This make it impossible for everyone to have vehicles. He says that smart fleets will make it easier for people to remain mobile. He believes that cars will be able to communicate with each other and that roads will be safer.
Lopez and his comrades are trying to keep their focus on making Synthia more capable of deciphering different data.