Traceability refers to the ability to trace at any time how the requirements placed on a system have been implemented, simulated and validated and which artifacts are associated with which requirements. Although this is not a new topic, the trend towards smart products means that it has grown enormously in importance. Electronics and software control an ever increasing number of safety-critical functions, not only in automobiles but also in other products, and these functions need to be validated virtually. One example would be autonomous vehicles. Testing every imaginable driving situation on the road would be far too time-consuming and risky. If their vehicles are to be certified for use on the road, carmakers and their system suppliers must, for example, be able to provide detailed proof of which situations they have simulated under which circumstance and with which tool chains.
Traceability is not only an issue for companies in the context of functional safety and compliance with the associated traceability requirements; it also plays a key role when it comes to digitalizing their business processes. Companies in every industry are trying to design their processes more consistently and therefore have to cope with a growing flood of digital information. The real challenge is not managing the digital information but rather managing the relationships and dependencies between the individual information objects. It is, for example, impossible to reliably assess the impact of changes without this knowledge about relationships.
Especially in complex development projects, e.g. in the shipbuilding industry and the mechanical and plant engineering sectors, project participants from different disciplines and domains today need to expend a great deal of communication effort to determine which data and documents correspond to the current development status – and all the more so since they often belong to different organizations. At certain milestones or quality gates, they have to prepare and collate the deliverables in what is a largely manual process in order to gain an overview of the progress being made in a project and possible deviations from the planning status.
Traceability is becoming increasingly important in the context of the digital twin and support for new, data-driven business models through digital twin applications. It establishes the link between the digital master or digital thread and the digital representative of the product that is actually delivered and provides the basis for enabling information from the operating phase to be fed back into and reflected in product development. Without this link, it is impossible – or takes a great deal of effort – to trace errors that occur during operation back to their possible roots in the development process, analyze them and thus eliminate them faster.
There are therefore many good reasons to explore the question of how to ensure traceability as efficiently as possible. Traceability is an essential prerequisite for providing evidence of compliance with the relevant standards and maturity models. It ensures greater transparency in the interdisciplinary product development process and plays a key role in speeding up product development and improving competitiveness thanks to innovative services. This benefits not only product developers, project managers, quality managers and service technicians, but ultimately also a company's partners and customers.
Today traceability is made more difficult by the fact that the different disciplines and domains create and manage their information objects and development artifacts using hundreds of different IT systems, which are often only integrated in a rudimentary form. System landscapes are also changing very dynamically because new products call for new and better technologies for their development and production. Therefore a key requirement for any solution for ensuring traceability is that it functions independently of the IT systems used.
It is our opinion that traceability can no longer be ensured using conventional integration approaches – at least not with an acceptable level of effort. Instead of replicating the relevant data in a higher-level system, the approach that we are pursuing involves the lightweight linking of information objects located in different source systems. What is crucial here is that we do not generate the links retrospectively but rather determine from the start which information objects are to be related to each other and how, while at the same time taking account of the relevant standards and maturity models. This is the fundamental difference between our approach and other linking concepts. Please contact us if you would like to learn more.