TU-TESL

• 22

According to foreign media reports, before the popularity of cars, vehicles had been equipped with sophisticated computers. More and more Internet vehicles need 150 million lines of code, far more than 145000 lines of code for the Apollo 11 moon landing program in 1969. Self driving cars may require up to 1 billion lines of code. For automakers, passengers and repair shops, cars that run on software rather than machines represent an extremely complex world of mobile travel, and it is not enough to just check vehicles, and brakes to identify faults. In order to enhance people's trust in the new generation of vehicles, Aurora labs, a start-up company based in Tel Aviv, Israel, has developed a "self healing" software, an active remote system, which can detect and repair potential vehicle faults, and automatically update and verify vehicle software without stopping the vehicle. "The automotive industry is facing the biggest change so far," said ori Lederman, co-founder and chief operating officer of Aurora labs. "Before all complex technologies and software emerge in the car, the most important thing is whether people trust such software and technology, otherwise they can't completely transfer the automatic driving right to the car, which also brings many challenges to the car manufacturers." Lederman added that new challenges include whether software problems can be detected after the vehicle is sold, whether the problem can be solved safely and reliably, and whether the problem can be solved without interrupting vehicle use. In 2018, in the United States alone, about 8 million cars were recalled due to software defects. Roger Ordman, executive vice president of marketing at Aurora labs, said: "before you faint, the human body can detect something wrong, the automatic immune system will indicate that there is a problem, and what can be done to fix it: raise your temperature or the number of white blood cells. Sometimes, the body will repair itself, sometimes not enough, and external intervention is needed. Our technology also adopts the same method. Before catastrophic failure, we detect whether there is a problem, point out exactly where the problem is, take measures to solve the problem, and make the system run smoothly and smoothly. " The company's line of code behavior technology, supported by machine learning and artificial intelligence technology, can deeply understand what software is installed on more than 100 vehicle engine control units (ECUs) and the relationship between software. In addition to the ability to detect software failures, the technology also allows the vehicle to update software remotely and wirelessly without stopping. Ordman added that, similar to the automatic update of smartphone applications, automakers are also able to continuously update and improve the software running on connected vehicles. However, automakers need to comply with more stringent regulations, including those relating to cybersecurity and "wireless" updates that are braked by agencies such as UNECE. Without the need for a large-scale recall of vehicles, Aurora Labs' self-healing software is able to use short-term repair functions to ensure that the functions continue to operate, and then the vehicle can be fully upgraded. Aurora labs, founded in 2016 by Lederman and Zohar fox, CEO, has raised $11.5 million (about 80.6277 million yuan) in multiple rounds of financing. Currently, the company is working with some of the world's leading automotive industry players, including major automakers in Germany, the United States, South Korea and Japan. (all pictures in this article are from Aurora labs)