AdVitam Architecture

This project can be divided into 5 Work Packages (WP) which are in continious interaction.

[WP1] Project Management and Dissemination

This Work Package has a twofold objective:

♦ Assure proper project management and coordinate the collaboration among all the partners
♦ Disseminate the results of the project to the scientific community, to bachelor and master students as well as to the society in general.

Both management and dissemination activities will be carried all along the project.


[WP2] Psychophysiological model of the driver

The assessment of the driver state is very important to adapt and tailor the interaction between the driver and the vehicle in the shared control model. The goal of this Work Package is to assess several psychophysiological indicators (e.g. ECG, EDA and EEG) and eye  movements in experimentally manipulated special driving situations in a driving simulator. Special driving situations include monotonous driving, driving under sleepiness and fatigue, noise (e.g., distractions from family members), and high attentional load (e.g., fog and limited visibility, high traffic density, speeding).
In addition, the same indicators will be assessed in the field, with test drivers driving a real car. Advanced signal processing techniques will be applied to the various measures in order to model continuously and in real-time the driver’s fitness to drive. In this regard, intelligent multimodal fusion of brain and peripheral signals will be  investigated, with the expectation to increase the accurate detection of the driver’s fitness to drive.
An additional crucial objective of this WP is the definition of a common evaluation framework for
the work of the three PhD students. The evaluation framework will define a set of driving scenarios with different levels of difficulty with respect to supervision and intervention tasks. The framework will be used during the work of the students to compare and assess the effectiveness of different interface solutions and alert messages as well as of the developed psychophysiological model.


[WP3] Human-Vehicle Interaction for Intervention

This Work Package is focused on the Human-Vehicle Interaction aspects involving the driver intervention on the system. In particular, we can highlight three main objectives:

♦ Design of the guidance system to help the physical interaction with vehicles
♦ Design of multimodal interaction for warning messages
♦ Dynamically adapt the interactions taking into account the context and the driver state


[WP4] Human-Vehicle Interaction for Supervision

The objective of this Work Package is to investigate advanced interaction modalities based on fullbody and multisensory experiences aiming at improving safety and user experience in semiautonomous vehicles.
In particular, WP4 aims at investigating how human-vehicle interfaces can be designed to be operated at different levels of attention according to the Interaction Attention Continuum (Bakker & Niemantsverdriet, 2016) with a special focus on peripheral interaction to support supervision and situational awareness. Wearable and ubiquitous computing will work as technological enablers of novel interaction modalities that are more natural and, additionally, can reduce the cognitive workload on the user.


[WP5] AdVitam final demonstrator

This Work Package has a threefold aim:

♦ Development of virtual environments to be used during the tests with the simulator.
♦ Converge the work of WP2, WP3 and WP4 into a global model for Human-Vehicle Interaction in shared control. As presented in Figure 4, data acquisition solutions developed in WP2 and WP3, will allow feeding the psycho-physiological model conceived in WP2 in real time. Consequently, the interaction solutions conceived in WP3 and WP4 will be able to dynamically adapt to the user needs and state.
♦ Provide a technological framework to integrate the different solutions developed during the project in a modular way. This task will involve the definition of a shared development and test environment as well as a demonstrator integrating the most valuable components.
The software platform will be developed in order to share data inside the research partners and to external researchers via the creation of a scientific challenge.