Options
2018
Journal Article
Titel
Functional reliability of cognitive control systems for manufacturing processes
Abstract
Industrial automation has led to a significant gain in reliability and stability in manufacturing processes through the introduction of numeric control and sensors, thus enabling control loops. This development has been identified as the third industrial revolution, which has been researched extensively over the last decades and led to an almost complete makeover of the industrial sector. Control loops typically feature a fixed goal value as well as control parameters, thus limiting the abilities of the machine in case deviations from the originally specified environment and application occur. As products are being more and more customized and their lifecycles have been shortened dramatically, classical control loops for manufacturing processes need to be enhanced with more flexibility and finally autonomy to meet these challenges. Self-optimization or the enhancement of control loops with cognitive capabilities have been identified as one way to achieve this flexibility: these systems are able to identify their own current status as well as the environment conditions and can deduct control strategies accordingly. Typically, they are enhanced with the ability to learn to enable working in yet unknown future conditions. On the other hand, such systems will only be successful if safety and security as basic requirements of smart factories can be ensured. Safety features many different aspects, with functional reliability being one of its most prominent. This contribution thus researches the functional reliability of cognitive control systems for manufacturing processes. For that an exemplary reliability model is developed using fault tree analysis. The model is evaluated by applying it to a validation case for force control in a turning application. It is shown that this modeling approach can be used to evaluate functional reliability in cognitive control systems.