The use of rFpro’s highly-accurate virtual environment will significantly accelerate the development of autonomous vehicles
Hertfordshire, UK, 18th September 2018... Global computer technology company NVIDIA has announced rFpro as a software partner at the GPU Technology Conference Japan 2018. rFpro will provide the virtual environment used to develop autonomous vehicles (AV) in NVIDIA DRIVE™ Sim and NVIDIA DRIVE Constellation.
NVIDIA DRIVE Sim is a platform that simulates a self-driving vehicle’s sensors – including cameras, LiDAR and radar. It drives the company’s NVIDIA DRIVE Constellation, a new hardware-in-the-loop virtual reality AV simulator. rFpro’s photorealistic environment ensures that training and development of the systems accurately translates into the real world.
“In order to precisely evaluate their safety, virtual development is the only realistic way to subject AVs to the uncountable number of scenarios that could take place on the road,” says Chris Hoyle, rFpro Technical Director. “It would simply take too much time to gather the data required. Simulation has become essential to the development of autonomous vehicles. Best of all, nobody can get hurt in a virtual world!”
At the conference, NVIDIA CEO Jensen Huang commented in his keynote speech, “We have to simulate everything. Simulation is the foundation of good engineering. Simulating artificial intelligence systems is vital to success.”
A ‘3D Reconstruction’ process allows rFpro to create ‘Digital Twins’ of the real world, which ensures simulated testing matches the real-world testing. It allows the vehicle developer to physically model their vehicle’s interaction with the world, through its sensors (cameras, LiDAR, radar) and through its vehicle dynamics.
rFpro employs a technique called Physical Simulation (illustrated in this video https://youtu.be/6UepFgkhcDg), which means that lighting, weather, the atmosphere, even the way the detailed road surface is fed to the customer’s virtual-vehicle are based on the laws of physics.
“Our Physical modelling is unique as it ensures that simulation can be validated and certified, so that the process can become part of the regulatory approval process,” explains Hoyle. “Taking into account vehicle dynamics also means that a driverless vehicle can be engineered, from day one, for customer acceptance. This is a key roadblock to the adoption of autonomous vehicles – customers must actually want to ride in them, or the billions invested will be wasted.”
rFpro, a UK business, founded in 2007, has been delivering simulation solutions to the Autonomous industry since 2015. There are autonomous vehicles testing today at off-road proving grounds, such as this one at Millbrook (https://youtu.be/xXOYseIJIRA), that were developed, tested and trained in rFpro.
NVIDIA’s AV products are being used by vehicle manufacturers such as Toyota, Mercedes-Benz, Volvo and Subaru as well as major Tier1s, such as Denso, Bosch and Continental.
“Driverless vehicles are expected to save hundreds of thousands of lives each year, because of their ability to be more safe than human drivers,” concludes Hoyle. “If rFpro and NVIDIA can help to bring this day forward, by just one year, there is the potential to save a huge number of lives.”
rFpro provides driving simulation software, and 3D content, for Deep Learning Autonomous Driving, ADAS and Vehicle Dynamics testing and validation. rFpro is being used to train, test and validate Deep Learning systems for ADAS and Autonomous applications by OEMs and Tier-1s. When developing systems based on machine learning from sensor feeds, such as camera, LiDAR and radar feeds, the quality of the 3D environment model is very important. The more accurate the virtual world is the greater the correlation will be when progressing to real-world testing. rFpro’s HiDef models are built around a graphics engine that includes a physically modeled atmosphere, weather and lighting, as well as physically modeled materials for every object in the scene. 100s of kilometers of public road models are available off-the-shelf, from rFpro, spanning North America, Asia and Europe, including multi-lane highways, urban, rural, mountain routes, all copied faithfully from the real world. rFpro scales from a desktop workstation to a massively parallel real-time test environment connecting to customers' autonomous driver models and human test drivers.
+44 (0)1295 277 050