My true passion   -   product design

system design

During my 13 years in R&D and especially in my role as head of a mechanical design group I had several times the chance to do a system concept for complete new products or big sub systems. In the beginning my team and I deployed the three steps of concept design more by intuition but later by a more and more formal way.

1. the development of alternative solutions for each requirement
2. evaluation of all alternative solution
3. selection of best combination for the complete system

The over time more systematic approach was supported by a growing number of applied development methods: formal patent and competition analysis, abstract modeling of system functions (basic function, block, flow diagram) creativity techniques, morphologic box, value analysis, quality function deployment, failure tree analysis and others.

Despite all methodical support I still believe that experience together with a deep and wide engineering knowledge is most important for designing complex systems successfully and efficiently. A system designer must have a detailed understanding of numerous physical correlations of the complete technical system and their subsystems in very different aspects (heat flow, dynamic behavior, acoustics, …). They need to have the ability to do both exact calculation of the system behavior and a quick rough estimation of the influence of system parameter to the total performance. Intuition and gut feeling are good accelerators and help to avoid try and error loops. In order to create a robust system concept suitable for serial production the system designer must know additionally about the mayor disturbance variables of real life and the typical variances of a real manufacturing process. 

I think for no other development task it less possible to replace talented and experienced engineers by just employing more but less gifted engineers. On the other hand it's also obvious that nobody can know everything in all technical aspects. Creating a culture for open minded (and talented) people and active discussion on one hand and a strict early and very frequent design reviews by independent (not in project involved) and experienced engineers are probable the best methods for system design.

mechanical concept design

Prior to the detailed mechanical designing is the definition of the mechanical concept in which subassemblies, modules, main components and the their assembly methods are defined. At least for serial products and even more for multi variant serial products it is important to stay away from 3D CAD and to keep the abstraction level high. For multi variant serial products it may also be a good idea to define a modular concept with a high reuse of components.

I always put a special focus on the definition of a robust assembly by reduction resp. control of variances in the assembly process. All assembly operation must be designed to deal with the manufacturing tolerances and be executed defined and by good control of the worker (e.g. fatigue-proof handling, good accessible joints, feedback effects, use of function tested subassembly, application of Poka Yoke principles like key-lock elements. positioning elements, easy distinguishably part and part orientation). Based on the first mechanical concept a first concept for in process quality control (test points) and for serviceability can be done as well - also without the need for 3D modeling or a long bill of material.

mechanical design

3D-modellierung, creation of drawings and bill of materials, tolerance setting, creation of delivery instructions, supplier negotiation, design of packaging, product documentation, the design of sheet metal parts, welding assemblies, milling and turning parts, parts made of stainless steel, aluminum, plastic injection molding, silicone extrusion profiles, rapid prototyping parts, high temperature plastics, …   -  All this has been the center of action during 5 years as head of mechanical design.

Cost minimization by manufacturing optimization must always be a strong focus of all design actions. This means both optimizing the design to the manufacturing site or country specific (standard) manufacturing capabilities and optimizing the internal and external manufacturing capabilities to the typical part spectrum. 

Cost minimization requires a certain greedy mindset, fun to be ungenerous. I have learned that it is easier to reach the quality and performance goals from a low cost design than the other way round. A bit exaggerated can be said: from cost perspective and quality perspective only a not existing part is a good part. It can't fail and costs nothing. I think a cost efficient robust design can only be achieved when parts, modules and the whole device is designed without a lot of reserves and which is improved if needed but then consequently. Once you've added an extra screw to your design it is very likely that this screw will be build in to end of life of the product a couple of thousands times. Everything, which is right for serial products, is wrong for small serial products and one-piece designs.

Good knowledge about design methods and about typical manufacturing techniques is mandatory for every designer. But due to the progress in technology nobody can know the limits of all processes and due the complexity of the global market nobody can know the regional availability and costs. Frequent and early design reviews internal, with supplier or with independent external consultants are again powerful methods.

electrical & electronic design

By education and role these are not my area of expertise. But over the years I could learn a lot and witness quite a bit of mistakes and typical traps from non certified cables over EMC problems and solutions to object-oriented modeling with UML (Unified Modeling Language).

Robust Systems - our services for product design