Wayve tests autonomous cars in new cities
Wayve, the London-based company pioneering autonomous driving vehicles, has tested its cars in new cities across the UK.
Wayve’s approach to self-driving cars is different to anything else on the market and that’s why the Virgin Group invested in the company. Although several technology companies and vehicle manufacturers have invested billions of dollars into self-driving technology, Wayve focuses more on machine learning and artificial intelligence, rather than engineering. Wayve is pushing the forefront of machine learning into the world of autonomous driving vehicles, learning to drive from data and experience.
Wayve has taken its self-driving cars on the road and into new UK cities to test how well they perform in cities that they haven’t experienced before. The goal was to see if its AV2.0 model that was trained in London could generalise its driving intelligence to new cities, without being given data on those destinations.
The trials took place over a three-week period in September 2021 in Cambridge, Coventry, Leeds, Liverpool and Manchester. Wayve selected cities and routes that varied in terms of road layout, road features (such as traffic lights, pedestrian crossings and bus lanes), driving complexity, and traffic density. For example, the areas in Manchester had more pedestrians, bus lanes and traffic lights, whereas the Leeds testing routes had almost four times as many cycle lanes.
The test provided promising results that Wayve’s AV2.0 model can in fact generalise its learned driving skill to new places. Without any city-specific adaptations, the autonomous driving system learns to drive more than 610km in previously unseen cities, demonstrating the skills it learned in London.
So what’s next? This multi-city generalisation test is a stepping stone to Wayve’s goal of bringing autonomous driving to 100 cities. It demonstrates what AV2.0 can offer in terms of simplicity, efficiency and scalability to bring self-driving vehicles to new, previously unseen places. And it also shows that teaching autonomous vehicles to drive in London can fast-track deployment anywhere.
Wayve will continue to develop its self-driving vehicles, teaching them to drive across different routes, urban environments and roads to make sure that they are well experienced in all kinds of driving situations.
Wayve has also been working on a Fleet Data Collection Pilot with delivery company DPD. The two businesses are exploring how innovations in computer vision and machine learning can be applied to existing fleet and delivery operations to increase the safety of smart urban delivery solutions.
“Real-world driving data is fundamental to building the core capabilities of Wayve's technology and we have built industry-leading expertise in the collection and utilisation of fleet-scale data,” said Alex Kendall, Wayve CEO. “Working with DPD is an incredible opportunity to accelerate the collection of peta-byte scale datasets that expand our coverage in more areas of the UK and helps us improve the safety and driving intelligence of our technology.”
Data collection devices have already been deployed on 50 DPD vans in Greater London. This enables Wayve to collect data from the vans during their normal operations. Wayve’s camera-first system works off 4G connectivity and provides a 360-degree view of the vehicle without impacting the driver.
Visit Wayve to learn more about how they are building the technology to make autonomous vehicles a reality.