'Design for all': methods and data to support designers

If designers are to meet the needs of the growing population of older and disabled people then data on size, shape, posture and capabilities will be increasingly important. This thesis details a methodology for the collection of anthropometry, joint constraints, reach range, postural capability and task specific information, to create a unique database of `individuals'. These data were then used in the development of a computer-based design tool (HADRIAN), to allow design professionals to estimate the percentage of people who could be accommodated by a design. Having complete data sets for individuals is vital to enable multivariate analysis, as opposed to traditional univariate percentile data. Following a review of the literature two interview surveys were conducted with 32 design professionals and 50 older and disabled people. The majority of designers were aware of the philosophy of `design for all', but rarely considered the approach due to perceived time and financial costs. With respect to older and disabled people it was found that nearly all experienced problems completing basic activities of daily life, and that improvements to existing designs could improve quality of life. Activities such as being able to cook a meal, and use the bath were reported as being particularly important. Firstly, a pilot study was conducted with 8 participants to assess the different data collection options. Data were then collected on 100 people, with the majority being older and/or disabled, and encompassing a wide range of capabilities. From these data it was possible to see that the anthropometric data showed a range beyond 15` and 99`h percentile for each dimension when compared to existing anthropometry data, and a breadth of variation in task specific behaviours. Validation trials were then conducted to compare the actual task performance of 10 of the 100 `individuals' with that predicted by HADRIAN, with postures and task capabilities being correctly predicted for open-access reach-and-lift tasks. This gives some confidence that it is possible to predict postures and capabilities from the data collected.