Autonomous Driving: A Convoy Function (Leading partner: AVL)

The use case setting includes a lead and follower vehicles, all of which rely on exact functional operation and user acceptance based on human perception of safety. The perception is the decisive factor for driver’s acceptance and ability to take over control from the autonomous driving system. The system functions seek cooperation between autonomous (safety-critical) CPSoS, IoT and vehicle’s advanced control strategies. The cooperation also relies on: accurate positioning of the ego-vehicle and the leading vehicle; velocity profile and planned trajectory of the leading vehicle; information on human behaviour for user acceptance and transition scenario in Level 3 / Level 4 autonomous driving, where human needs to be in the position to take back control over the vehicle. The combination of the above functional interactions stresses the challenge of creating a reliable solution for autonomous and safety-critical distributed systems of networking computing elements and humans in the autonomous driving context. The aspects of energy-efficiency, dependability, cyber-security and support for multi-criticality are hence of crucial importance.


A heterogeneous, composable and scalable architecture for CPSoS applications will be defined and realized. The proposed solution levers on HW diversity, as a relevant paradigm for safety critical architecture, through state-of-the-art automotive multi-core microcontrollers, from Infineon and NXP, featuring different characteristics in terms of safety mechanisms and computational structures. The CPS integrates also an FPGA, thus providing further flexibility and processing capabilities. Furthermore, the CPS features enhanced connectivity through HW encryption thanks to HSM technology embedded both on state-of-the-art microcontroller and on smart transceivers. It is worthy to note that the proposed CPS is an open HW/SW platform supporting the development of complex algorithms for a wide range of autonomous and safety-critical applications (Obj. 1). The design methodology will support the analysis of such an AI based control system under a cyberattack, exploiting the functionalities made available by AI modules embedded at the edge. In this context, AI is a crucial enabler of personalization. Moreover, the proposed CPSoS will also feature enhanced connectivity through encryption (Obj. 2) (HSM technology on state-of-the-art microcontroller and smart drivers). The use case demonstrates TEACHING’s innovation aspect through human integration into the loop (Obj. 6). Monitoring of biological changes and their association to the users’ emotional status (e.g. electrodermal activity, electromyography, electrocardiography, postural and acoustic features, etc..) provides valuable information for development of AI algorithms (Obj. 7) that reinforce the dependability of the CPSoS. In particular, one of the main biological activities that will be monitored is the cardiovascular activity, as any measure based around hearth and blood vessels is potentially correlated to a wide range of information regarding different physical and psychological conditions. The monitoring of blood flow in peripheral vessels can be achieved using a range of technologies (e.g. plethysmography, IR camera, etc). A combination of sensors can be successfully used for the analysis of physiological parameters (e.g. camera, odour, body movement, etc..) so that the CPS will efficiently take care of sensors data fusion through AI methods.