Transceiver manufacturer Aquantia, technology company Bosch, manufacturer of tyres and braking systems, among others Continental, developer of graphics cards NVIDIA and car manufacturer Volkswagen establish the Autonomous Vehicles (NAV) Alliance. This new alliance aims to develop and standardise high-speed network connectivity in cars, something the parties involved say is crucial to enable level 5 autonomous driving.
A large number of companies are working on autonomous vehicles. Such vehicles are equipped with sensors and continuously collect large amounts of data. This data is used, among other things, to determine the exact position of the vehicle, determine the optimal route to a particular destination and detect obstacles. The amount of data collected by autonomous vehicles is only expected to increase further as vehicles offer higher levels of autonomy.
'Key to level 5 automation'
"Redundancy and various AI algorithms are key to level 5 automation (ed: where vehicles can drive fully autonomous in any scenario). The amount of data collected by different types of sensors (camera, radar, lidar, ultrasound) can quickly reach 32TB every eight hours. This level of data transfer requires a new kind of superfast networks, such as multi-gig Ethernet. The NAV Alliance is going to catalyse the development of reliable next-generation networking platforms for self-driving cars," explains James Hodgson, Senior Analyst Autonomous Driving at ABI Research.
The parties involved point out that an autonomous vehicle is essentially a data centre on wheels. This data centre functions as an "advanced nervous system" for self-driving vehicles. The nervous system consists of ECUs, CPUs, GPUs, high-definition cameras, sensors, gateways and storage devices. This equipment is connected via a high-speed Ethernet network, enabling speeds of several gigabits per second (Gbit/s). This nervous system is called 'Networking for Autonomous Vehicles' (NAV) by the parties.
Objectives
The NAV Alliance has several objectives:
- Developing an ecosystem for next-generation Multi-Gig Ethernet networks for use in autonomous vehicles.
- Establishing specifications for in-vehicle network interoperability, security and reliability.
- Promoting products and solutions that meet the new specifications.
- Partnering with standardisation organisations.
- On setting up marketing activities to raise awareness about network architecture for autonomous vehicles.
"The realisation of fully autonomous driving and the vision of an accident-free future depends significantly on seamless communication within vehicles. To help meet the level 4/5 requirements of the NAV Alliance's development and testing work, Bosch supplies vehicle computers, central gateways, high-resolution cameras and sensors. Together with our alliance partners, we are helping to create a standardised network that is both safe, reliable and redundant and helps bring autonomous driving to our roads," said Viktor Molnar, Senior Vice President Body Electronics at Robert Bosch.
'Multi-Gig Ethernet has already proven itself'
"Autonomous vehicles require an onboard AI (ed: artificial intelligence) supercomputer designed for functional safety and capable of processing large amounts of data from sensors on redundant and diverse deep neural networks and algorithms. Multi-Gig Ethernet has proven itself in terms of interoperability and scalability, making it a logical choice for automotive connectivity and delivering critical data from the sensor suite to the vehicle's AI brain," said Gary Hicok, Senior Vice President of Hardware Development at NVIDIA.
Interested parties can join the NAV Alliance. More information about the alliance and opportunities to join can be found at NAV-Alliance.org.
Car Connectivity Consortium
Incidentally, the NAV Alliance is separate from the Car Connectivity Consortium, despite its domain name NAV-Alliance.com suggesting otherwise. The Car Connectivity Consortium focuses on connecting smartphones and vehicles, using the MirrorLink standard to control smartphones from the dashboard or steering wheel.
Author: Wouter Hoeffnagel
Source: NAV Alliance
Source: Aquantia
Source photo: Pixabay / 0532-2008
