Automobile interiors are increasingly designed with the aid of three-dimensional computer-rendered representations of the human body, but the use of these models is hampered by the lack of accurate posture prediction. This research develops and evaluates statistical models to predict static driving posture and proposes a general theory of driving posture selection based on a muscle-activity-reduction criterion.

The effects of vehicle and seat geometry on driving posture were examined in reconfigurable vehicle mockups and vehicles. The postures of 84 drivers spanning a wide range of body size were recorded in the mockups for various combinations of seat height, steering wheel position, seat cushion angle, and downward vision restriction. The postures of 120 men and women were recorded in five vehicles after a 15-minute drive. Drivers adapt to changes in the vehicle and seat geometry primarily by changing limb posture. Each of the test variables has significant effects on driving posture that are independent of anthropometry.

Three alternative whole-body posture prediction models developed from the laboratory data were assessed by comparing model-predicted driver eye locations with observed in-vehicle eye locations. The average eye-location error for the model that combines direct prediction of eye and hip locations with inverse kinematics is less than 4 mm horizontally and less than 7 mm vertically. Prediction of mean postures for population groups is considerably more accurate than the prediction of individuals' postures.

In an electromyographic study, the average driving posture chosen by 10 men was generally consistent with the hypothesis that drivers tend to select torso postures with the highest eye location that can be achieved with low levels of thoracolumbar extensor muscle activity. Biomechanical simulations demonstrated that passive lumbar flexion stiffness may contribute significantly to the stability of the upper body in driving postures.

This research quantified the effects the driving environment on whole-body driving posture and developed new insight into torso posture selection. The findings improve the utility of human models for vehicle ergonomic assessment and contribute to an improved understanding of human adaptation to task environments.