Mission Profiles

In the field of electronic systems, the mission profile has been one of the key concepts since the start of the scientific examination of the subject of reliability. Its exact meaning varies with time and the industry using it. In particular, over the course of increasing digitalization and networking in the context of IoT and the opportunities resulting from this, the subject of mission profiles is once again becoming the focus of research and consideration is being given to innovative approaches.

But what exactly is a mission profile, what is it used for and how has this concept developed? Due to constant change and different perspectives the exact content is hard to define, but ultimately it is a combination of influences that an electronic product is subjected to in its specific application during its service life, and which have an effect on its reliability. This may involve both environmental factors, such as temperature or vibration, as well as performance parameters such as current or voltage.

The loads can be represented in different ways. A qualitative representation has been available in the very well-known MIL standard since 1962. It describes the mission profile with rough allocations to the site, ranging from “ground” with the additions “benign,” “fixed” or “mobile” to “naval” and rocket launch, and launch from a cannon. In this case, each of these categories is assigned what is known as a π factor, with which the fault rate of a component determined for standard conditions is multiplied. The π factors are empirical parameters. They include both aging effects and the constant fault rate due to random failures. This representation already very efficiently fulfills the purpose of comparing the fault rates of different components and making estimating their applicability under various conditions possible. However, it is not suitable for determining the lifetime of a single element.

Quantitative considerations are made possible by means of aging models. These investigate the influence of a single stress variable on the service life of a component. To do this, all failure mechanisms that may be responsible for a fault must be considered and tested as to how they respond to changes in a stress variable. For example, aging processes often run faster at an elevated temperature. In this case, it is possible to deduce changes in the service life at different temperatures using the Arrhenius equation. An advantage of this approach is the direct quantifiability of the load and the at least theoretical transferability between different components, as long as the underlying physical failure mechanism remains the same.

However, as a stress variable is virtually never constant in real applications, the principle of accumulated damage is applied. It is assumed that the temporal sequence of loads is irrelevant, and that only the total period that is spent in a stress state is decisive for aging. Thus, all times that an object has spent under a similar load can be added up, even if it was exposed to other stresses in between. Mission profiles can now be estimated relatively simply from known application data. Monitoring approaches to record real, individual mission profiles are also very easily possible by means of this, since only the temperature must be measured and few accumulating time values ​​must be stored.

Monitoring the stress variables that affect an individual electronic system becomes particularly attractive, however, if the time sequence of the stress variables is also to be included in the service life prediction. This makes it possible to consider recovery effects. For example, it may occur that an electronic system that spends the first half of its life at a low temperature and is then operated continuously at an elevated temperature fails significantly earlier than a system in which the two states alternate several times. On the other hand, the aging may also be accelerated by a frequent load change, which usually happens by stimulating an additional failure mechanism.

Complex aging models that determine the remaining service life of an electronic system from a variety of recorded variables and, in addition to this, take into account not only their variables but also their chronological order are currently not yet fully developed. Often, the overlapping of various physical aging mechanisms is too complex to make satisfactory reproduction in the laboratory possible, thus allowing the effects to be separated and individually modeled. For this reason, in recent years more and more monitoring approaches have been used, which record a large amount of data and correlate this data purely statistically with failures. However, this approach, which is not dissimilar to qualitative considerations, is no longer based on empirical values ​​but works with the help of AI and big data. For the future, the research field is seeking to develop so-called hybrid models that will combine the benefits of physical modeling and data-driven prediction of failure probability.

By this combination of classic methodology with the latest data processing methods, mission profiles will continue to retain their central role in the reliability assessment of electronic systems, even if they are continuously changing in terms of content.