Research of fast and efficient solutions to a technical challenge always requires a suitable methodology when creating innovative products. The way we proceed to discover efficient solutions is based on three methods:
Analytical Model: an analytical model often reveals important parameters underlying a problem. It is therefore a powerful mean to find creative technical solutions but it is unfortunately restricted to simple cases. When the analytical model becomes too complex, the solution is then to use the analytical model only for a part of the problem under analysis or to switch to the second method.
Computer Simulations: computer simulations are often considered as the panacea when creating new products. The advantage of computer simulations is that they can give a vision of the product and its behaviour before being prototyped. Computer simulations also give access to hidden variables that are not measurable on a prototype. The drawback of computer simulations is that the computer time required to get useful results quickly becomes huge.
Experimental Results: when inventing a new product, building prototypes of one or several subsystems can quickly give useful pieces of information about its behaviour.
The efficient approach we suggest is to use the above tools when they are optimum in the course of the research and development steps. The right tool is the tool that gives the most information in the shortest time frame. By continuously keeping in mind the above three tools and by alternately selecting them, product development can be very fast and lead to numerous new ideas.
At each iteration, knowledge increases and opens the door to new ideas.
To be concrete, let's start with an example: the design of an innovative 10kW power supply topology.
A new circuit can be imagined by crossing the product requirements, the knowledge of the engineers and by parsing a lot of literature in order to avoid re-inventing the wheel. Once it is done, the analytical model is understood but not verified. A computer simulation can be useful to check that the circuit works as expected. If necessary, the analytical model can be corrected. When confident about the result, a prototype can be elaborated. When bringing-up the prototype, the computer simulation model can be refined for the next product development.