Technology companies Bosch and Nvidia announced plans to jointly develop artificial intelligence self-driving systems for mass market cars.
The firms are developing an AI self-driving car computer built on Nvidia’s deep learning software and hardware that enables vehicles to be trained on the complexities of driving, operated autonomously and updated over the air with new features and capabilities.
“Using DRIVE PX AI car computer, Bosch will build automotive-grade systems for the mass production of autonomous cars,” said Nvidia’s chief executive Jen-Hsun Huang. “Together, we will realise a future where autonomous vehicles make mobility safe and accessible to all.”
Nvidia has also revealed that it is developing an autonomous truck system with with global manufacturer Paccar.
The truckmaker, which manufactures Kenworth, Peterbilt and Daf, has developed a proof-of-concept self-driving truck with SAE Level 4 capability built on Nvidia DRIVE PX 2 technology, trained on deep neural networks. Level 4 capability is defined as a system that drives itself.
“Paccar is exploring automated driving systems and we are excited about what our collaboration on artificial intelligence with Nvidia has delivered so far,” said chief executive Ron Armstrong.
The potential size of the market for self-driving trucks is huge: there are currently 300 million trucks worldwide, driving more than 1.2 trillion miles annually.
Autonomous acquisitions
Intel, which expects the driverless market to be worth as much as $70 billion by 2030, made its first move into the sector this month with the acquisition of Israeli autonomous vehicle technology company Mobileye for $15.3 billion.
Mobileye – which accounts for 70% of the global market for advanced driver-assistance and anti-collision systems – offers a range of technology and services including sensor fusion, mapping, front- and rear-facing cameras.
In January, Mobileye announced it was developing a test fleet of autonomous cars with BMW and Intel. Meanwhile, Tesla began incorporating its technology into Model S cars in 2015.
"Mobileye brings the industry's best automotive-grade computer vision and strong momentum with automakers and suppliers," said Intel chief executive Brian Krzanich.
"Together, we can accelerate the future of autonomous driving with improved performance in a cloud-to-car solution at a lower cost for automakers."
Mobileye has used STMicroelectronics to produce chips which its sells to many of the world’s top car manufacturers for its current, third-generation of driver-assistance systems. However, while working with BMW, it teamed up with Intel for its fifth-generation of chips that aim to be used in fully-autonomous vehicles.
Earlier this month, BMW said it will leverage its partnership with Intel and Mobileye to release a self-driving car in 2021, which will be able to handle most driving tasks but still requires a driver to intervene.
However, whether this happens depends on regulations, the availability of 5G telecommunications and whether there is the infrastructure in cities on which autonomous cars can travel.
Australian initiatives
Meanwhile, across the other side of the world, an Australian state is set to begin autonomous vehicle trials for its airport and one of its public universities.
The South Australian government has announced AU$5.6 million (£3.5 million) in funding for driverless shuttle and pod trials in Adelaide. This includes $1 million for Adelaide Airport to trial three electric driverless shuttles operating between the long-term parking lot and the terminal. An additional $1 million will go to Flinders University, which will collaborate with the Royal Automobile Association of South Australia on a three-year $4 million driverless shuttle project to transfer students from public transport to its Bedford Park campus.
Another $1.8 million has been awarded to international driverless car supplier RDM Group to begin transporting goods at Adelaide’s Tonsley innovation precinct, with the aim of developing a market-ready autonomous delivery pod within a year.
The compact size and agility of the shuttles allow them to operate at an increased frequency, potentially operating 24 hours a day and will help to significantly lower carbon emissions.
The announcement came as Australia’s first driverless bus trial, the Intellibus in Perth, notched up its 2,000th passenger. The NAVYA bus, which began trials in August 2016, uses LIDAR, cameras, GPS and odometry to provide 3D perception that enables it to map the environment, detect obstacles and interpret traffic signs. The bus, which can transport up to 15 passengers at a time, always has someone onboard who can take the controls if necessary.
With the potential for millions of autonomous vehicles to be on the road in just a few years, there’s a lot at stake, so we can expect company developments to continue to come thick and fast in coming months.