The UK will not gain the full safety benefits of self-driving cars until every car on the road is connected to each other, according to IAM RoadSmart.
Pointing to a new report produced by the University of Michigan, the motoring and road safety charity says until that point is reached, the human mind ‘holds the edge’.
The white paper, ‘Sensor Fusion: A Comparison of Sensing Capabilities of Human Drivers and Highly Automated Vehicles’, concludes that computers are ‘generally well suited to perform tasks like driving’ – especially in regard to reaction time, power output and control, consistency, and multichannel information processing.
However, the report, produced by Brandon Schoettle*, concludes that ‘human drivers still generally maintain an advantage in terms of reasoning, perception, and sensing when driving’.
In the report, Mr Schottle says: “While no single sensor completely equals human sensing capabilities, some offer capabilities not possible for a human driver.
“In the short to medium term AV [autonomous vehicle] sensing systems will still be critical for detection of any road user or roadway obstacle that is not detected and shared by connected vehicles which is where the human brain comes in.”
He also points to a number of circumstances in which both human capability and a connected vehicle’s perception can be compromised – thus increasing the need for each party to work together – such as extreme weather, excessive dirt or physical obstructions, darkness or low illumination, large physical obstructions and dense traffic.
The report also highlights areas where the human brain ‘wins out’ over a vehicle’s ‘brain’, arguing that for functions including memory, reasoning, sensing and perception, human involvement is both desirable and advantageous.
In March, IAM RoadSmart warned that cars with growing levels of autonomy could make motorists lazy and over reliant on gadgets – with ‘far reaching implications’ for the potential reduction of people killed and seriously injured on the roads.
Neil Greig, IAM RoadSmart director of policy and research, said: “The ultimate win-win situation is a place where information from each vehicle is shared with the vehicles around it; add that to human experience born from a lifetime of ‘trial and error’ and you have the ideal double-act to spot crashes before they happen.”
*Brandon Schoettle is a project manager at the University of Michigan Transport Research Institute’s (UMTRI’s) Human Factors Group. Since joining UMTRI in 2000, Mr. Schoettle has authored over 100 scientific articles and technical reports.
His recent research with the Sustainable Worldwide Transportation (SWT) program focuses on international road-safety trends, changes in driver licensing and demographics, eco-driving strategies and monthly monitoring of new-vehicle fuel economy and emissions.
Category: Autonomous vehicles.
I can see the logic of the IAM statement, and that makes a lot of sense. But, I live in Somerset and can you imagine this happening with tractors, agricultural vehicles in general and (very often) crusty old Landrovers and the like, most of which are more likely to see sheep than other vehicles.
Nigel ALBRIGHT
0
Research by Ford has led them to the conclusion that either you have a vehicle driven by a human, with collision avoidance technology for when they get it wrong* (or circumstances change to make previous decisions wrong). Or you have a vehicle driven solely by a electronic chauffeur – taxi driver – with no human driving interface. The hinterland of a vehicle that can drive itself but then calls for help is impracticable as the time for a driver to become aware and take full control exceeds two seconds, by which time any undesirable event will probably have occurred or passed.
*wrong can include accidentally exceeding the speed limit.
Mark, Cardiff
0