Data Farming Touted as AV Simulation Breakthrough
A UK automotive graphics specialist is claiming a breakthrough to cut the costs of autonomous vehicle testing using ‘data farming’.
The company, rFpro, claims to have developed a means to slash the hardware costs associated with large-scale AV simulation, removing the industry’s dependence on time-consuming and error-prone manual annotation of test data that is created frame-by-frame.
The company says its data farming process is similar to render farming used to drastically reduce the costs of animation film making. It allows complete datasets to be built covering the full vehicle system where every sensor is simulated and synchronized at the same time. This is essential when employing sensor fusion to bring together data, for example from multiple 8K HDR stereo cameras, LiDAR and radar sensors at the same time.
The new approach claims to allow testers to start with a single computer to perform a complex simulation involving multiple sensors. Matt Daley, rFpro managing director, said: “This new approach from rFpro provides a digital, cost-effective way of creating the same data completely error-free and 10,000 times quicker compared to manual annotation, which takes around 30 minutes per frame with a 10% error rate. This step-change will enable deep learning to fulfill its potential because it significantly reduces the cost and time of generating useful training data.”
— Paul Myles is a seasoned automotive journalist based in London. Follow him on Twitter @Paulmyles_
