The skill of bicycle riding
The principal theories of human motor skill are compared. Disagreements between them centre around the exact details of the feedback loops used for control. In order to throw some light on this problem a commonplace skill was analysed using computer techniques to both record and model the movement. Bicycle riding was chosen as an example because it places strict constraints on the freedom of the rider's actions and consequently allows a fairly simple model to be used. Given these constraints a faithful record of the delicate balancing movements of the handlebar must also be a record of the rider's actions in controlling the machine. An instrument pack, fitted with gyroscopic sensors and a handlebar potentiometer, recorded the roll, yaw and steering angle changes during free riding in digital form on a microcomputer disc. A discrete step computer model of the rider and machine was used to compare the output characteristic of various control systems with that of the experimental subjects. Since the normal bicycle design gives a measure of automatic stability it is not possible to tell how much of the handlebar movement is due to the rider and how much to the machine. Consequently a bicycle was constructed in which the gyroscopic and castor stability were removed. In order to reduce the number of sensory contributions the subjects were blindfolded. The recordings showed that the-basic method of control was a combination of a continuous delayed repeat of the roll angle rate in the handle-bar channel, with short intermittent ballistic acceleration inputs to control angle of lean and consequently direction., A review of the relevant, literature leads to the conclusion that the proposed control system is consistent with current physiological knowledge. Finally the bicycle control system discovered in the experiments is related to the theories of motor skills discussed in the second chapter.