A study by the UK’s University of Southampton has shown the length of time needed for a driver to switch from automated vehicle control to manual control is crucial for the safety of future driverless vehicles.
Simulations run by a team at the university showed a very broad range of ‘control transition times’ for participants to resume control of their car. The researchers believe their findings will be important for system designers when considering the lead time needed to take control of a vehicle, and suggest the focus should not just be on the average time needed for a person to successfully switch, but rather on the range of resumption times.
Professors Neville Stanton (right) and Alexander Eriksson (below) found that, under non-critical conditions, drivers needed between 1.9 and 25.7 seconds to take control from automation. Such a large range reflects a variety of driver behavior and environmental conditions.
The study was carried out using the Southampton University Driving Simulator (SUDS), which comprises a Jaguar XJ saloon vehicle linked to the STISIM Drive M500W System, with comprehensive vehicle dynamics model and active steering. SUDS is an interactive driving simulator with three driving displays and 135º driver field-of-view, using in-vehicle driving controls with high-resolution digital sensors and speed-sensitive steering feel, and STISIM Drive simulation software. The simulated road environment includes all road types (urban, rural, dual carriageway, motorway) and other road users (dogs, pedestrians, cyclists, motorcyclists, cars and lorries). The road users can be scripted to produce any behavior on demand.
The authors observed 26 men and women (aged between 20 and 52) engaged in simulated driving at 70 mph, with and without a potentially distracting non-driving secondary task. They recorded response times as the drivers took over or relinquished control of the automated system.
A takeover request was issued at random intervals ranging from 30 to 45 seconds during normal motorway-driving conditions. The authors found that drivers engaged in a secondary task, prior to a control transition, took longer to respond, posing a safety hazard.
The researchers warn that if the lead time for normal non-critical control transitions is based upon data obtained in studies using critical situations, there is a risk of unwanted consequences.
“We hope our findings can guide policymakers in setting guidelines for how much lead time a driver will need when changing in and out of automation,” Stanton commented. “The challenge for designers is accommodating the full range of response times rather than limiting parameters to mean or median transition times.”
Eriksson explained, “Too short a lead time, for example seven seconds prior to taking control, as found in some studies of critical response time, could prevent drivers from responding optimally. This results in a stressed transition process, whereby drivers may accidentally swerve, make sudden lane changes, or brake harshly. Such actions are acceptable in safety-critical scenarios when drivers may have to avoid a crash, but could pose a safety hazard for other road users in non-critical situations.”
