Intelligent robots will progressively take over the role of humans in complicated or dangerous tasks on the road, in industry or in hospitals. Future autonomous systems will therefore increasingly use AI-based components to improve performance. Safety, autonomy, and reliability play a particularly important role. To enable widespread use in these safety-critical environments, they must therefore be developed according to high quality standards.
For this reason, FOCETA brought in the last three years leading academic research groups and industry partners together to advance the foundations for the continuous engineering of trustworthy, adaptive autonomous systems. The project adopted an integrated approach that combines the advantages of data-driven and model-based techniques. The FOCETA concept was implemented in open source tools and tested in demanding and difficult application scenarios such as urban transport automation and intelligent medical devices. This approach demonstrated its viability, scalability and robustness while addressing the European industry's needs for cutting-edge technology.
The technical foundations and assumptions on which the traditional principles of safety technology are based and which have arisen as part of continuous development have not previously been applied to autonomous systems capable of learning. The results from FOCETA and the solution approach developed have shown that continuous engineering can better manage the challenges in safety-critical scenarios throughout the entire product life cycle.
The fortiss team contributed significantly to the success of the project by developing methods for system-level testing, using search-based approaches developed by fortiss researcher Lev Sorokin to find security-related edge cases. In addition, his team colleagues Tewodros Beyene and Radouane Bouchekir Tewodros Beyene and Radouane Bouchekir from the Software Dependability competence field have constructed a rigorous and scalable distributed framework for generating safety and security workflows in order to continuously curate appropriate reliability evidence for FOCETA's industrial use cases.
The FOCETA consortium consisted of the following high-profile partners from research and industry: University of Grenoble Alpes, DENSO, Intel, RGB Medical Devices, Siemens, University of Liverpool, Bar-Ilan University, Austrian Institute of Technology (AIT), Institute of Applied Information Processing and Communications (aka IAIK) at the Technical University of Graz, Aristotle University of Thessaloniki and the L-UP SAS Paris.