The CPSoSaware architecture is going to be tested in two different pilot sites (Germany and Italy) by performing trail scenarios to two different use cases.
The first use case is focused on connected semiautonomous vehicles where we will perform trails focused on Human in the loop scenarios, like non predictable failures that may involve the human driver and how this affects the design operation continuum support of the CPSoSaware solution as well as human situational awareness enhancement when using the CPSoSaware architecture. We also use this use-case to access the cybersecurity mitigation strategies using the CPSoSaware architecture and its response to cyberattacks.
Use case 1 Connected and Autonomous Vehicles
One of our goals is to evaluate the CPSoS system in the automotive domain, considering different connected cars use cases, for validating issues related human in the loop control, reliability and security.
We will consider scenarios in which Autonomous Driving Systems (ADS), that are able to work unattended only under mild conditions, while they require a human driver to take control in situations that cannot be handled in an automatic way by issuing a so-called Request to Intervene (i.e. SAE level 3 vehicles). Such systems require the cooperation of a Driver State Monitoring System (DSM), that assesses the state of the human driver (e.g. by performing pose-estimation, or emotion recognition, possibly utilizing multiple modalities) and a sub-system that performs an analysis of the scene outside the vehicle and controls the vehicle to move autonomously on a predefined path (e.g. recognizing vehicles ahead, estimating their velocity/trajectory, forecasting future vehicle locations). In the concept of CPSoSaware, driver monitoring will not only be used to assess the driver alertness but also to recognize the driver and check his/ her level of capacity to safely drive the vehicle. Other applications for the CPSoSaware system will also involve operations related to gaze and facial expressions analysis to assess the driver’s mental state. These developments severely contribute to ensure higher safety.
drowsiness will be modelled upon the three different symptoms that it produces:
(1) Vigilance, which can be described as a state of watchfulness and a process of paying close and continuous attention to detect certain unexpected little changes in the environment.
(2) Fatigue mainly derives from performing a highly demanding task for extensive time periods (“time-on task”). It can be defined as weariness or exhaustion from labor resulting in a feeling of stress, an aversion to further exertion and the feeling to be unable to carry on with the task.
(3) Sleepiness is the state of being ready to fall asleep.
We distinguish drowsiness measures into two (2) broad categories:
The second use case will be focused on HRC in the manufacturing environment and will involve trails that challenge the MOOD CPSoSaware concept and trails on accidents/failures as well as cybersecurity attacks that challenge the collaborative control mechanism and the autonomic decentralized operation of the CPSoSaware solution as well as the Design operation continuum support in the presence of cybersecurity attacks. The two use cases complement one another since they have different requirements and specificities (open spaces and moving CPS, close interconnection with humans versus closed space environment, static CPS and more relaxed interaction with humans).
Use case 2. Human – Robot Interaction in Manufacturing Environment
The concept developed by CPSoSaware can be applicable in complex CPSoSs that need to be operate under highly demanding conditions (e.g real time response time, safety) like industrial manufacturing domain. However, in such systems the complexity of their design and management makes it very hard to cope with for the systems’ operators as well as human users. The proposed inclusion of AI assistance that is used in CPSoSaware as well as the inclusion of the human factor in the overall approach makes the solution very attractive for industry. The CSoSaware solution has the capability to include and update the description of the complex model by updating the scope and dependencies of the sub-systems in case a modification in one or more sub-systems occurs. CRF will utilize the CPSoSaware solution in a complete industrial manufacturing use case that involve collaborative robots and human operators/users as well a associated logistics system. Whenever an industrial manufacturing CPSoS of this kind includes actors such as robots or automation machines, there is a need for the CPS programming code and control loop process to be updated during CPSoS operation phase following a redesign trigger due to changes in the manufacturing value chain. This redesign is a complex issue since it must be in coherence with all the other sub-systems including humans (through their Human Machine Interface (HMI)). In a complex manufacturing hierarchy in future manufacturing environment, the human operator will have a central role in the factory itself. New collaborative robots, as well as complex logistic systems have to consider the operator as a “virtual agent” or as a sub-system in the CPSoS description in order to deal the proper flux of the information in the factory.