Next-generation smart products will require a next-generation engineering process. Each cycle of new product development will inevitably be more complex. Products will continue to evolve from purely mechanical devices into a combination of mechanical parts, electronics, with a growing number of controllers, and software they will become mechatronic. Traditional mechanical engineering processes will need to evolve, as they do not support the optimization of such mechatronic systems.
Bringing worlds together
The more features a product requires to remain competitive, the greater the control burden becomes. Design-Test-Fix has long been replaced with a simulation-driven process based on new product geometry and enabled by CAD and CAE technologies.
Mechatronic products present a new challenge; the mechanical, electronics and programming design worlds must be brought together.
Early attempts at a systems engineering approach have not been brought together in a way that proves viable for the pace of most new product development today and will certainly not be adequate for the future.
We need to develop a new level of understanding, an evolution from rapid prototyping, using physical hardware, to a model-based engineering approach, based on existing ‘plant’ models, using simulation models representing the controlled systems.
Competition will not allow manufacturers to eliminate complexity by deleting product attributes. And if you launch immature products to the market and you may face expensive recalls in the short term and long term brand reputation damage.
The traditional engineering process is under scrutiny after major problems leading to program delays, cost overruns and even cancellations, with examples in every industry. More often than not, the issues lie with unanticipated interactions between elements that are only discovered during integration, testing and, worst case, during service.
Model-Based Systems Engineering
The risk of high-profile program failures in every industry can be avoided.
The problem is that most products in most industries are based on the continual development and refinement of well-established concepts using a familiar, linear, process. Fundamental product and process innovation is rare in most companies but it is exactly what is required: a fundamental change to an innovative, new, Model-Based Systems Engineering process.
Model-Based Systems Engineering, MBSE, involves decomposing a design into separable elements, characterizing the intended relationships, and verifying the system is built and operates as intended.
‘Models’ will offer contextual information on different degrees of freedom and interactions. Fully understanding the dynamic behavior of every interaction of every implemented system is crucial.
By analyzing and designing the ideal architecture upfront, even before the concept phase, engineers can manage scarce time better and minimize risks during the development process. This assumes the availability of high fidelity models (or plant models) for control model development (MIL) and control software development (SIL) as well as for validation of the actual controllers (HIL).
A call for a profound innovation
New, more robust, engineering platforms should enable verification without the fabrication of several prototypes.
If we want to leap frog the competition in time and cost savings, we need to re-design the end-to-end development process. By re-defining traditional test engineering, the 1D system and 3D performance simulation approach, integrating and taking this new framework of Model-Based Systems Engineering as the starting point, we will realize quantum leaps.
The evolution towards integrating platforms of modularized subsystems and components is already going on. The need to frontload and calibrate controls and software to balance upfront conﬂicting attributes for simultaneous mechanical, thermal, electronic and controls development is, to some extent, yesterday’s news. Tomorrow’s solution is the new model-based-framework.
Engineers now have to integrate traditional test and simulation engineering into this platform, and include domain specific data management that federates with new model-based-systems-level PLM-backbones.
But more important, they should also strive for an holistic open-innovation approach with extended interfaces for the full supply-chain and, preferably, the whole eco-system. The way manufacturers and their suppliers work together will fundamentally change too. The sharing of engineering data requires a new respect for intellectual property and a new approach to warranty and in-service issues.