Deterministic Design: Schedules 


Although there are many different elaborate models of how designs develop, most would agree that overall there are essentially three basic phases to the development of products: Strategy & Concept, Detailed Engineering & Development, and Integration & Test. Each of these phases, when executed properly, take about a third of the total development time. Too much or too little time spent on one of the phases means that you will likely be rushed in one of the other phases, and quality and performance will suffer. 


There are many different detailed methodologies ranging from House of Quality1 to Rapid Product Development2 to Axiomatic Design3; however, all of them require a sequencing of events and thus all require the engineer to maintain a schedule. Needless to say, there are an equally large number of scheduling methods, but at this phase, the method is not as important as the net effect. What matters is that the total time allotted for completion of the project be partitioned to each of the basic phases, and then each phase be further partitioned to ensure that an appropriate amount of time is allotted to each task. A simple table often thus suffices to outline what needs to be done, which by the way, is the same process used to manage the creation of a major written work, or the composition of a symphony, or the cooking of a large meal, or creation of a major software program! The patterns of best design practice repeat! 


Given a project that is already broken up into thirds, the best way to decide how time should be spent in each of the phases is to assess the cost/performance issues associated with the project. Starting with a fundamental assessment of the physics of the problem and of possible solutions, one is likely to identify potential leading edge strategies, and possibly even a disruptive technology. These potential strategies should be investigated first, and then a detailed development schedule can then be created. 


Each of the thirds can be thirdified into thirdlets! Given a 12 week development process for a robot to be entered into a design contest, the first trimester is allotted to developing strategies and concepts. The first week should be devoted to developing a deep understanding of the problem and using this understanding to create strategies ranging from shoot-from-the-hip instinct ideas, to ideas driven and guided by analysis. The second week should be spent doing experiments and more detailed calculations to weed out lame or super risky ideas, and evolving a leading edge strategy. The third week is then spent generating, creating, and evolving concepts to arrive at a best concept for development. The fourth week is float time, It might be needed because more than expected experiments and analysis was required during the third week; or a leading edge concept might indeed have been identified, and then the next trimester can be entered early and you can get ahead of schedule! 


The second trimester is for detailed engineering and development. The first two weeks are for dividing the concept up into modules and for developing and testing the most risky module. Accordingly, at the end of the sixth week of the project, the most risky module should actually be all designed and ideally a Bench Level Prototype completed and tested. The next two weeks are for completing the detailed engineering of the remaining modules. If there is any uncertainty, a module can be engineered and a Bench Level Prototype can also be designed and tested. 


The first two weeks of the third trimester are for completion and testing of remaining modules. The eleventh week is for integration and testing of the modules, and the twelfth week is for continual use to find any flaws that may remain. If all goes according to schedule, and it will if one is careful to follow a deterministic design process, the entire process of conceiving of an idea, growing it, and then giving birth to it will be a joyous happy experience! 


Remember, “It takes 9 months to gestate a human baby, no matter how many women are assigned the task!”4


Carefully read through ALL of the milestones required for the development of your project and develop a strong internal urge and commitment to work deterministically and to adhere to the schedule. Start off on the right foot by completing Milestone ONE as soon as possible! Engineers who consistently miss milestones will soon find themselves looking for burger flipping  


1.K. Otto, K. Wood, Product Design, Prentice Hall, Upper Saddle River, NJ, USA 2001

2.Karl T. Ulrich, Steveen D. Eppinger Product Design And Development, 2000 The McGraw-Hill Companies, Inc. Boston, MA, USA

3.Nam. P. Suh, Axiomatic Design, Advances and Applications, 2001 Oxford University Press, New York

4.Frederick P. Brooks Jr. The Mythical Man-Month, 1995 Addison-Wesley Longman, Inc. Reading, MA. USA






本网站由阿里云提供云计算及安全服务 Powered by CloudDream