Approaches to efficiency in manufacturing and assem bl y over the past 100 years have concentrated frequently on the twin themes of div iding jobs done by people into defined segments and to automating as many of tho ...Approaches to efficiency in manufacturing and assem bl y over the past 100 years have concentrated frequently on the twin themes of div iding jobs done by people into defined segments and to automating as many of tho se segments as possible. This approach to manufacturing engineering may be char acterised as one of deconstruction and replacement so far as the role of the hum an being is concerned. More recently design engineering has been approached in a similar light with research into ’intelligent’ methods involving, for example, expert systems or genetic algorithms to handle an increasing proportion of desi gn tasks. Expert systems are predicated on the idea that it is possible to capture and systemise significant amounts of professional knowledge: a task that has proved with experience to be much more difficult than when such systems wer e first proposed. Genetic algorithms, notwithstanding considerable research eff ort, have yet to make significant inroads into the practise of design and manufa cture. This paper will draw on new and previous experimental work carried out with supp ort from industry to propose an alternative approach that is enabled by the prod ucts of modern information technology. In this paradigm it is recognised that t he involvement of the human design expert is, for the time being at least, inevi table. Rather than seeking to reduce or even eliminate the " expert" with his or her tacit knowledge, the aim is to provide the expert with interactive to ols including immersive virtual reality so that implicit, human expertise can be applied effectively in the design cycle. The proposed approach is considered in the context of a particularly costly and difficult task, namely the design and planning for manufacture and assembly of c able harnesses for use in electro-mechanical artefacts. The efficient and reli able manufacture of cabling systems for many such products in the aerospace, aut omotive and IT sectors provides diverse challenges to designers, manufacturers a nd cable installers. Cable harness layouts are often so complex that design tend s to be carried out as an end activity, which may lead to higher costs, or even a product redesign. Current practice usually requires scaled or full-sized phys ical prototypes onto which the cable layout is constructed as the last productio n process. This is the case even when advanced CAD systems are used for this an d other parts of the design. Problems encountered at the cable harness design st age have a marked impact on the time needed for new product introductions. A prototype immersive system for cable harness design, developed by the research team has been used tested in experiments carried out in collaboration with indu stry. This work has shown that the task completion times obtained by the partici pants were between two and five times faster in than current CAD systems; even t hough the experiments were heavily biased towards CAD. The Virtual Reality inter face, with fewer mouse clicks and keyboard inputs, appears to have a significant impact on completion times. Cable harness design requires in-depth three-dimensional spatial reasoning whi ch CAD systems tend to lack due to their conventional flat screen display. Resul ts so far show evidence that the two or three degrees of freedom of movement on a flat screen provided by current CAD cable harness systems inhibits the user’s ability to route cable layouts especially in complex 3D assemblies. Immersive V R, on the other hand gives users the ability to change their viewpoint quickly a nd provides six degrees of freedom of movement thus showing great promise as a s uitable environment for routing cable. Initial results and feedback from the industrial partners are encouraging and su pport the theme of the research paradigm. Rather than seeking to reduce or even eliminate the ’expert’ with his or her tacit knowledge, the aim is to provide t he expert with interactive tools including immersive virtual reality so that implicit, human expertise can be applied most展开更多
文摘Approaches to efficiency in manufacturing and assem bl y over the past 100 years have concentrated frequently on the twin themes of div iding jobs done by people into defined segments and to automating as many of tho se segments as possible. This approach to manufacturing engineering may be char acterised as one of deconstruction and replacement so far as the role of the hum an being is concerned. More recently design engineering has been approached in a similar light with research into ’intelligent’ methods involving, for example, expert systems or genetic algorithms to handle an increasing proportion of desi gn tasks. Expert systems are predicated on the idea that it is possible to capture and systemise significant amounts of professional knowledge: a task that has proved with experience to be much more difficult than when such systems wer e first proposed. Genetic algorithms, notwithstanding considerable research eff ort, have yet to make significant inroads into the practise of design and manufa cture. This paper will draw on new and previous experimental work carried out with supp ort from industry to propose an alternative approach that is enabled by the prod ucts of modern information technology. In this paradigm it is recognised that t he involvement of the human design expert is, for the time being at least, inevi table. Rather than seeking to reduce or even eliminate the " expert" with his or her tacit knowledge, the aim is to provide the expert with interactive to ols including immersive virtual reality so that implicit, human expertise can be applied effectively in the design cycle. The proposed approach is considered in the context of a particularly costly and difficult task, namely the design and planning for manufacture and assembly of c able harnesses for use in electro-mechanical artefacts. The efficient and reli able manufacture of cabling systems for many such products in the aerospace, aut omotive and IT sectors provides diverse challenges to designers, manufacturers a nd cable installers. Cable harness layouts are often so complex that design tend s to be carried out as an end activity, which may lead to higher costs, or even a product redesign. Current practice usually requires scaled or full-sized phys ical prototypes onto which the cable layout is constructed as the last productio n process. This is the case even when advanced CAD systems are used for this an d other parts of the design. Problems encountered at the cable harness design st age have a marked impact on the time needed for new product introductions. A prototype immersive system for cable harness design, developed by the research team has been used tested in experiments carried out in collaboration with indu stry. This work has shown that the task completion times obtained by the partici pants were between two and five times faster in than current CAD systems; even t hough the experiments were heavily biased towards CAD. The Virtual Reality inter face, with fewer mouse clicks and keyboard inputs, appears to have a significant impact on completion times. Cable harness design requires in-depth three-dimensional spatial reasoning whi ch CAD systems tend to lack due to their conventional flat screen display. Resul ts so far show evidence that the two or three degrees of freedom of movement on a flat screen provided by current CAD cable harness systems inhibits the user’s ability to route cable layouts especially in complex 3D assemblies. Immersive V R, on the other hand gives users the ability to change their viewpoint quickly a nd provides six degrees of freedom of movement thus showing great promise as a s uitable environment for routing cable. Initial results and feedback from the industrial partners are encouraging and su pport the theme of the research paradigm. Rather than seeking to reduce or even eliminate the ’expert’ with his or her tacit knowledge, the aim is to provide t he expert with interactive tools including immersive virtual reality so that implicit, human expertise can be applied most
