Engineering Design Performance

Extended Abstract A reader may be slightly puzzled by the title. Indeed, it might say either what performance in engineering design is, or how to engineer design performance. We mean both. Research we are going to report about aims at developing a rigorous engineering methodology for reaching optimal performance in the processes of engineering design. The methodology and the software tool are based on a fine-grained model of performance and its environment. The domain we specifically focus on for validation, evaluation and further exploitation is engineering design in microelectronics and integrated circuits. However, the models presented in this talk are more generic and may be applied in various industrial sectors or academic disciplines. Our research in formalizing design performance has solid industrial motivation. In the sector of microelectronic and integrated circuits design this motivation can be demonstrated using a system law formulated by Gordon Moore [1]. One of its important corollaries says that the number of components on a chip doubles roughly every 10 to 24 months. It is known in the sector that doubling the number of components crammed onto a silicon chip causes approximately the order of magnitude increase in engineering design effort [2]. In contrast, the productivity of engineering design environments (electronic design automation tools, technologies, methodologies, skills), although increasing, grows substantially slower. Since time to market frame remains roughly constant, the only way to fit this window is hiring more designers and blowing up the budgets. Given a fixed design environment, ensuring that an engineering design process is performed in an optimal way may bring serious benefits, lowering the mentioned gap between design productivity growth and the increase in chip complexity. Engineering design performance should be measured in a way keeping the process on the most optimal possible trajectory. Today's performance measurement and management practices are based mostly on strategic level benchmarking – finding a place of a company among the others on the industry sector bench. The prevailing methodology is the use of balanced scorecards [3]. Such a benchmarking provides reasonably sound indications of what is good or bad in terms of performance at a strategic management level. However, it does not fully satisfy industrial demand. Krause [4] has analyzed the following weaknesses of the contemporary business performance management (PM) approaches pointing to them as to the reasons of dissatisfaction in industry: (i) strategic PM approaches are