Simulator-based Driver Training: Re-inventing the Wheel?
24th May 2013
With a lack of appropriate research data available, the MOD sponsored a new research programme to provide evidence-based data: the Training Experimentation Simulator Systems (Land) programme, or TESSL. TESSL was devised to discover whether simulation can be used to train handling and control skills for tracked vehicles, whether it is more appropriate for certain types of skills or tasks than others, and to what extent the method of training affects the motivation, engagement and confidence of trainees.
TESSL was a joint government/industry initiative: the Defense Science & Technology Laboratory (DSTL) designed a controlled experiment to compare live and simulator-based training effectiveness, while an industry team led by XPI Simulation supplied the research simulator and technical capability.
The challenge for the industry team (which also included CAE, NSC and Qinetiq) was to design and build a new simulator and simulated environment for training drivers of the Warrior tracked-vehicle – using mainly commercial off-the-shelf equipment and delivered within six months.
To build the simulated environment, XPI mapped the MOD’s established training range at Bovington. A highly-detailed scan of the range was achieved using the Light Detection and Ranging (LIDAR) technique, which can map the environment to an accuracy of 5cm.
The simulator consisted of the key items that would be needed in driving training: the seat, steering, pedals and other controls that a driver needs to understand and use; a driver’s eye view that is as realistic to the Bovington training range as possible, including the rear view generated in mirrors; and a simulated vehicle that responds as closely as possible to a real vehicle, including motion and sounds.
In January 2013, the TESSL programme then began a controlled experiment with 30 new drivers. After receiving their safety critical instruction together, half the group was trained using the simulator, and the other half on a real Warrior vehicle.
The two groups were then combined again and assessed on their ability to drive a real Warrior vehicle in set situations: a forward manoeuvre, a controlled brake, a depth perception test that requires the vehicle to be stopped between two points, and the ‘knife edge’ test, which requires the driver to balance the vehicle on the peak of a steeply-inclined 40-foot hill.
The assessments were rigorous, involving telemetry on the vehicle, time controls, task observations, and assessor rating. Crucially the assessors had no knowledge of how each driving had received their training.
Across all objective and subjective measures of driver performance, drivers who had been trained on the simulator performed as well as, or in some cases better, than drivers who had received live training.
The two groups had the same pass rate for the forward manoeuvre test, while drivers from the simulator group were slightly better at the chicane task. The difference was more marked in the knife edge test – in which the simulator group was significantly better.
Importantly, the feedback from the simulator group was largely positive; comments included that it was “fun to drive”, was “new, different and exciting”, and that it built more confidence because it allowed “mistakes which would be dangerous on the real vehicle.”
In providing substantive evidence on the use of simulators for training drivers of Armoured Fighting Vehicles (AFVs), the TESSL study has shown that it can both save money and improve the training experience and results.
This article is based on a joint presentation given by Simon Skinner, Managing Director of XPI Simulation, and Jessica Allen, Research Psychologist for the Defence Science & Technology Laboratory, at the ITEC conference, Rome, May 2013.