Autonomous vehicles – addressing potential risks
15 August 2016 • Author(s): Transport Research Laboratory (TRL)
Richard Cuerden, chief scientist, engineering & technology at the UK’s Transport Research Laboratory (TRL) discusses the potential issues surrounding the use of autonomous vehicles in smart cities following the latest Tesla crash in China…
“The latest Tesla autopilot collision in China further highlights potential issues around the use of automated systems, particularly in cases where the driver is still required to remain alert and attentive at the controls of the vehicle.
“The Society of Automotive Engineers currently specifies five levels of vehicle automation. Tesla’s autopilot system is classified as level two automation, which means the driver is required to maintain alertness and be ready at the controls, even in autopilot mode. This presents well-known challenges in terms of drivers’ awareness and understanding of the capabilities of the automation systems and the process by which control is shared and shifted between the driver and the vehicle in different modes of automated operation.
“We are going to see more collisions like this where, for whatever reason, the driver and the technology didn’t identify and react to the threat. What we need to do now is understand why the vehicle made the decisions it did and act accordingly. This is where projects like MOVE_UK. which will compare the behaviour of automated systems to human drivers, can really help. By understanding what went wrong and why, we can quickly teach collision avoidance systems to better predict any risks in real life environments.
“At the same time, it’s vital that drivers of vehicles with automated functionality remain aware and follow the instructions provided by the manufacturer, so that incidents like the one in China can be avoided as we discover more about this new technology.”