A machine that thinks for itself, learns and adapts

Cross-sectional project: Self-optimization

The reliability, user-friendliness and resource efficiency required from products and production systems are increasing as the quality and operating standards demanded by customers rise. At the same time, energy consumption must also be reduced in order to avoid high costs. There is great potential for improvement through self-optimization (SO) processes that integrate intelligent behavior into systems, allowing devices and machinery to adapt autonomously to changes in operating conditions. For example, self-optimizing energy management systems in electric vehicles can distribute the available energy based on the operating situation and with consideration toward conflicting objectives such as maximizing comfort versus maximizing range. This allows the available energy reserves to be used efficiently, achieving an optimum overall result.

The aim of the research project is to develop a set of tools that makes self-optimization methods and processes available in a user-oriented manner, so as to help companies integrate self-optimization into the mechanical engineering systems of tomorrow.

To that end, self-optimization methods and processes in the form of solution models are made available in a database. These include machine learning and cognition, intelligent control, regulation and data processing concepts, and mathematical optimization processes. The results are compiled to create a holistic development approach for self-optimizing systems. They are then validated by companies in the Leading-Edge Cluster and converted into marketable products and production systems such as self-optimizing industrial laundries, household appliances or manufacturing processes for production units and mechanical engineering.

The project makes a key contribution to the innovative leap from mechatronic systems to self-optimizing systems. Companies can considerably improve the resource efficiency, reliability and user-friendliness of their products and production systems, thus meeting market requirements and remaining competitive. The results are made accessible to other manufacturing companies – e.g. those involved in mechanical engineering, the electrical industry and the automotive supply industry – through transfer projects and disseminated outside the cluster by engineering firms. They are subsequently integrated into new university and further education programs.

Project duration
01 July 2012 - 30 June 2017