Hosted by IDTechEx
Artificial Intelligence Research
Posted on April 8, 2019 by

Learning to drive like a human

Humans have a remarkable ability to learn to drive quickly and can obtain a licence to drive across a whole country after tens of hours of practice. But after 10 years of commercial self-driving car development, over 10 million autonomous miles and $5B per year spent, we still do not have commercial self-driving vehicles on our roads. To turn what is currently a fantasy into a reality, UK startup Wayve decided on a different approach. For more information see the IDTechEx report on Electric Vehicles and Autonomous Vehicles in Mining 2018-2028.
The solution is machine learning, which is surpassing hand-engineered systems everywhere. Intelligent behaviour cannot be hand-coded, but can be learned through experience. Wayve have built a system which can drive like a human, using only cameras and a sat-nav. This is only possible with end-to-end machine learning.
  • No HD-Maps,
  • No expensive sensor/compute suite,
  • No hand-coded rules,
  • Driving on roads never-seen during training.
This scales self-driving technology like never before: for everyone, everywhere.
The video below shows the system driving on public roads in Cambridge, UK. It's driving on roads it has never been on before using just a simple sat-nav route map and basic cameras. The Wayve researchers don't tell the car how to drive with hand coded rules: everything is learned from data. This allows the vehicle to navigate complex, narrow urban European streets for the first time.
End-to-end deep learning
The Wayve system learns end-to-end with imitation learning and reinforcement learning to drive like a human, using computer vision to follow a route. Imitation learning allows the copying of behaviours of expert human drivers. Reinforcement learning allows learning from each safety driver intervention to improve the driving policy.
The model learns both lateral and longitudinal control (steering and acceleration) of the vehicle with end-to-end deep learning. The team propagates uncertainty throughout the model to allow them to learn features from the input data which are most relevant for control, making computation very efficient. In fact, everything operates on the equivalent of a modern laptop computer. This massively reduces the sensor & compute cost (and power requirements) to less than 10% of traditional approaches.
Source and top image: Wayve
Learn more at the next leading event on the topic: Business and Technology Insight Forums - Tokyo, September 2019 External Link on 18 - 19 Sep 2019 at Tokyo, Japan hosted by IDTechEx.
More IDTechEx Journals