Major progress towards the development of autonomous cars saw the Swedish transport authority in Transportstyrelsen approving self-driving tests on public roads in Sweden. The tests will be conducted in association with seasoned and trained drivers who will have their hands off the steering wheel with the cars allowed to move at a top speed of 80kmph.
The Swedish transport authority has defined three specific routes where the self-driving tests will be permissible. The three routes that have been approved are road 40 between Jonkoping and Gothenburg, E4 between Stockholm and Malmo, and E6 between Malmo and Gothenburg.
Dennis Nobelius, the Chief Executive Officer at Zenuity described the approval as a huge success towards the development of unsupervised cars in the country. Nobelius also revealed that conducting the tests would allow the company to collect important data and aid the organisation in improving their safety which would contribute to commercially rolling out autonomous vehicles.
Zenuity, a company that specializes in the development of software for autonomous vehicles and driver assistance technology is a joint venture between Volvo Cars and Veoneer. The company was founded in 2017 and currently employs over 600 employees in its offices in Sweden, USA, Germany, and China.
The company has been constantly working towards the development of autonomous driving systems. In October, the company launched a supercomputer specially designed to achieve high levels of automation in driving. The supercomputer titled Zeus uses data from radars, cameras, and other sensors to interpret a situation and follow a course of action. Zenuity claimed the supercomputer is capable of handling all the dynamic driving tasks with almost no human interaction required. The supercomputer Zeus leverages the scalable architecture of NVIDIA’s DRIVE AGX Xavier and is powered with the NVIDIA DRIVE OS. The NVIDIA architecture used in Zeus was manufactured by integrating six different processors that could seamlessly run diverse algorithms including the complex deep learning artificial intelligence algorithms. Leveraging the capabilities Zeus can process data from 27 different sensors at a time.