How can the Development of Embedded Systems be Optimized?

This is a question which we ask ourselves all the time. How can we get better results for our customers in a shorter time frame?

From this the engineering credo of Solcept evolved, on which our daily design work is based.

What are the Approaches for Optimization of Development?

The traditional approaches are structure-based, like TQM, process optimization and Six Sigma. These approaches lead to success in repetitive production, but are debilitating in innovation. This because innovative workflows are typically reactive, i.e. the next step is based on the result of the last which should be new and not foreseeable.

Several "lean" approaches for development try to solve this, in software development also the related agile approach.

The base of lean thinking are the two poles value and waste. The waste contains all the effort which does not generate value for the customer, so it should be minimized.

The development system at Toyota is considered to be furthest in this respect. It has been but on the following four pillars by the NCMS as "knowledege based development":

entrepreneurial-technical project leadership ("chief engineer")
expertise and experience
planning based on responsibility (bottom-up & key milestones/ gates)
set-based concurrent engineering

Of these the last point is especially interesting, because it says that the solution space should not be reduced already in the feasibility phase to one variant only. Instead several variants of components should be developed in parallel through several gates. The decision for or against a variant is taken if from e.g. prototypes enough information has been gathered.

This principle is know from agile development as "decide as late as possible but as early as needed".

It looks at first sight like waste. Considering that risky variants often produce changes late in the project, it can be shown that this method generates less time-to-market and effort than "fail and react". This is especially true if one of the variants is a proven known fall back solution.

The Pragmatic Approach of Solcept

The pragmatic approach of Solcept is the adaptation of the project management to the customer, the technology and the circumstances. We try to eliminate e.g. wasted documentation by keeping record of as many information as possible in the code or on the schematics. So no additional documents must be synchronized, administrated and consulted.

Thereby we draw upon the pillars of knowledge based development:

entrepreneurial-technical project leadership through senior designer as project leader
expertise and experience through excellent engineers and knowledge management
planning based on responsibility through planning by the designers themselves
set-based concurrent engineering through early problem solving

The last point is especially important for us and we try to evaluate, integrate and test key components as early as possible in order to mitigate as much risk as possible. This happens on one hand through virtualization (simulation) of complete systems including hardware, sensors and actors, on the other hand through emulation and early integration using rapid prototypes, mostly consisting of evaluation boards and partial breadboards.

This efficient hardware-software codesign results in an integration time for e.g. the final hard- and software in the range of a few days.

What is Our Experience?

In complex system projects we could already show fast time-to-market and very good cost control. However we try to follow another principle of lean thinking, which is the search for perfection, in order to optimize our procedures and projects further.

Profit from it in your next project: contact us, we gladly advise you.