The DATE2022 conference is intended as a meeting point for researchers, software and hardware designers, manufacturers of electronic circuitry, among others, to focus on technology and systems.
During this event the CARAMEL project participated in the discussion session dedicated to providing a platform for discussion of opportunities and collaborations for innovation and research in Europe, in which the project’s technical coordinator Peter Hofmann took part in the panel session “The Good, the Bad and the Trendy of Multi-Partner Research Projects in Europe“.
Although the conference initially had to be held face-to-face, it had to be reorganized in the short term towards an online format.
In the CARAMEL project, we strive to implement machine-learning based detection of attacks against the connected and/or autonomous vehicle by analysing sensor data, V2X data, and the status of embedded controllers like the OBU (on-board unit) in real-time using a tamper-proof device that directly integrated into the car – the anti-hacking device (AHD). This architecture and concept are novel and innovative, and CARAMEL is the first project to implement and demonstrate this.
Even as CARAMEL implements its concepts and architectures in the automotive space, the key CARAMEL innovations are transferable to other IoT and Embedded applications domains as well, such as factory floors, building automation systems or others. Therefore, the CARAMEL presentation and presence of CARAMEL representatives during the conference would be of value to the DATE 2022 community.
In the CARAMEL project, several integration options for the anti-hacking device based on different IoT devices have been pursued already. However, for the concept to be commercially viable and cost-effective, the concept of a machine-learning-based intrusion device must be even better integrated into commercial offerings for the Automotive and IoT market. The DATE 2022 community could provide valuable input for that endeavour.
CCAM and IoT both will face important security challenges in the future as bad actors discover these new areas for their activities. Since bad actors will use machine learning to subvert machine-learning-based processes and algorithms in the CCAM and IoT world (eg. using Generative Adversarial Networks (GAN)), a trend in the security industry is also to use machine learning to detect and counter these attacks. The CARAMEL project showcases this approach in the Automotive context.