Comparing Two Driving Simulation Scenarios for Steering and Speed Control

Abstract

A driving simulation experiment was conducted to test the effectiveness and relative efficiencies of two practice scenarios, each designed to present a series of lane changes for participants to negotiate. Scenario 1 included 50 lane changes and participants were instructed to follow the posted speed limits, which ranged from 48 km/h (30 mph) to 105 km/h (65 mph). Scenario 2 included 20 lane changes and participants began driving at 80 km/h (50 mph). They were told to increase their speed as they were comfortable to do so. In both scenarios, the lane changes were identified using sets of pylons flanking the desired travel lanes. Measures of steering and speed control were recorded. A cost function was developed as the product of the absolute value of the lane position of the simulator vehicle at the moment the vehicle reached each set of pylons and the square of the travel time between successive sets of pylons. The cost and cumulative cost per trial values were tested for whether they fit a power curve, indicating whether the participants learned, were still learning, or were not learning to interact with the simulator. Both scenarios were shown to be effective. The differences in the means and variances of the cumulative cost per trial values were tested. The results support the notion that Scenario 1 was more efficient, as indicated by lower variances after 20 lane changes. Additionally, the participants driving Scenario 2 had greater travel times between lane changes than those driving Scenario 1. The drawn conclusion is that directing participants to drive a range of speeds while making lane changes helps in their learning to interact with the driving simulator.