Visual demand in manual task performance - Towards a virtual evaluation

Doctoral thesis, 2006

Visual information is crucial for carrying out motor activities. However, knowledge about the effects of visual demand on the performance of manual tasks is not well documented. Visual determinants must be identified in order to be used in visualisation and simulation tools (e.g. computer manikins) as well as considered during ergonomic evaluations of workplaces and products. This thesis developed a visual demand model in which the characteristics of the task, the operator and the environment have an influence on the visual demand related to the performance of a task and result in a specific behaviour. This model has been used for the design of the five appended papers. The overall aim was to identify the determinants of visual demand in manual task performance. Two scenarios were investigated: car driving and manual assembly work. Paper I was an inductive study where the visual demand model was used to examine how the task, environment and operator characteristics influence the visual demand involved in assembly work. Paper II dealt with the effect of the task characteristic, e.g. precision, as determinants of visual demand on working posture and movement paths in a laboratory study. Paper III investigated the influence of the task and the environmental characteristics, i.e. time pressure, real motorway and interior car design as determinants of the visual demand on drivers’ visual behaviour and primary task performance. Paper IV examined the influence of the environment, the task and operator characteristics, i.e. time pressure, real motorway, interior car design, driver’s age and skill as determinants of the visual demand on a driver’s visual behaviour and primary task performance. Paper V studied how visual demand can be predicted using a visualisation tool such as a computer manikin. Numerous methods have been used to highlight the relation between visual demand and the behaviour produced by operators: an eye tracking system, motion tracking, observations, and interviews. Designing the studies with high level of validity was a priority and the majority of the studied focused on the subjects in their own environment, e.g. secondary task performance whilst driving in motorway traffic and assembly work in an industrial plant. The following visual determinants of manual task performance were identified: degree of precision degree required in manual assembly, fine-tuned activity in manual assembly, location of in-car controls, driver’s age and experience. Vision analysis carried out with computer manikin software does not really facilitate detection and consideration of visual demand, as it does not provide a method for analysing the visual determinants for a specific task. Suggestions concerning the design of the vision module and a method for how to use it are proposed. A further step will be to take into account psychomotor tasks and mental load into account in order to evaluate visual demand, i.e. study the influence of cognition on motor activity as well as its effect on task and workplace design.