Usability of User Interfaces of Computer-Aided Industrial Design Systems

Alexander Nikov, Roumen Jijanov
2000 Proceedings of the Human Factors and Ergonomics Society Annual Meeting  
The paper is aimed at studying the usability of user interfaces of computer-aided industrial design (CAID) systems and at the development of guidelines for their user interface design. For the determination of an aggregated evaluation using experimentally gathered data, the fuzzy knowledge-based evaluation (FKBE) algorithm (Stoeva, 1991) developed by one of the authors was employed. Based on fuzzy logic, it calculates this evaluation in a way similar to the human judgement process. An
more » ... ion of a CAID system was performed. The usability of its user interface in designing drawings of industrial design was measured and evaluated by 2 standard tasks, typical for industrial design: The first task consisted in a pure drawing of geometrical elements supplemented with hatchings on the 2D-level. The second task integrated complex operations on the 3D-level. For both tasks the quantity was almost the same with 2D and 3D geometrical elements. Thus a direct comparative evaluation of user interface usability was possible. The usability of the user interface was evaluated on the basis of expert scorings according to a checklist. The checklist consisted of 10 criteria including 125 checklist items. Based on the values of the checklist items, aggregated evaluations for 10 usability checklist criteria were calculated bottomup. For both tasks these values were determined by means of FKBE algorithm. Then top-down deficiency analysis was carried out and usability bottlenecks pointed out. We found significant differences in activity styles depending on the task. Time spend on 2D-task, drawing performance and design performance are better by up to 100% in comparison with 3D-task: number, extent, structure, syntax, and sequence of commands and parameters are more complex. Frequently additional information about the functionality of CAID-system had to be ob-tained to cope with the complexity of the system. Differences in usability can be assigned directly to the task, whereas the lower drawing performance in 3D-task is due to shortcomings of the CAID-system or to handling difficulties on the part of the operators. Differences in time spent on task and design performance of up to 100% show obvious shortcomings in the handling and support functions of the CAID-system. The same tendency can be diagnosed with respect to a comparison of 2D-and 3Dfunctionality. With increasing complexity in the design of the system, the interaction with a CAIDsystem becomes more complicated, the number of possibilities by which the designer can achieve a given goal increases, the work-task-related output increases, and the strain level required by the operators in order to achieve the result rises. Especially those functions which should reduce stress (routine work, e.g. dimensioning, hatching) induce a higher strain level. This effect can be explained by some functional deficiencies of the CAIDsystem, especially the complex use of the functions (large amount of parameters to be defined, restrictions and requirements for the use of a function, such as mathematically closed contours of areas for hatch, defined end-points of geometrical elements, etc.). For each of the usability criteria, relevant recommendations and guidelines for the design of user interfaces of CAID systems were developed. Their advantages, disadvantages and further developments are discussed in detail.
doi:10.1177/154193120004400220 fatcat:dcfn5jsqr5batnzlqojsxa5iqe