Home » Cross-cutting theme


Vehicule and mobility


VEM
The automobile creates problems related to road safety, energy consumption and pollution. These societal issues are a major concern for automotive manufacturers and the scientific community.



Keywords : Global vehicle automation behavioural and usage modelling traffic optimization.


Coordinator : Fabien CLAVEAU  

The considerable evolution of on-board technologies over the last twenty years has made vehicles safer, more comfortable and environmentally friendly. Driver assistance systems have developed, and it is the vehicle's range that is ultimately targeted. This autonomy requires the automation of the vehicle, which is based on improving perception through the fusion of various proprioceptive and external sensors, on trajectory generation, and on control-command which, from the target trajectory, seeks to determine in real time the commands to be applied to the vehicle's actuators.
The laboratory develops innovative embedded systems to make motor vehicles more intelligent. Intelligence here refers to the vehicle's ability to perform certain advanced functions independently, but also its ability to communicate harmoniously with the driver. In addition, many challenges related to the mobility of goods and merchandise depend on the use, integration and synchronization of vehicles in transport and distribution networks. The answers to these questions are based on the design, optimization and management of these networks. The LS2N is developing its action on these issues, focusing on two challenges :
  • Challenge 1 : Safe, intelligent, integrative, cooperative, autonomous or semi-autonomous vehicles: various driving assistance systems have emerged in recent years, ranging from low-level functions acting on vehicle dynamics (e.g. ESP) to the delegation of functions normally performed by the driver, in an increasingly integrated manner (ACC, parking assistance, pedestrian detection, etc.). Among the locks identified, we are particularly interested in the question of accelerating the development of these systems, which are becoming crucial given their proliferation and their application to a growing number of vehicles. Our contribution also aims to propose methodological tools supporting a global, structured, robust and efficient design approach, proceeding by the combined integration of different elementary control functions (braking, propulsion, steering, suspension...), with a view to the global automation of the vehicle (e.g. combined longitudinal/lateral control, global chassis control, etc.). In addition, the rise of autonomous vehicles raises many research questions about the place and role of the driver in the human-machine system. This includes, for example, monitoring the driver during the manual and autonomous driving phases, as well as studying the best transition modalities between these phases. These questions require the definition of driver models adapted to the current task. The laboratory is based on the experience acquired in the field of shared control between driver and PLC. Other related issues concern the perception of risk and the definition of acceptable trajectory envelopes for interaction with other users, as well as the design of new generations of multisensory and gestural interfaces.
  • Challenge 2 : Sustainable, interconnected, shared and mixed networks, synchronized, agile, reconfigurable, dynamic, robust, scalable: beyond the vehicle, the question of its use and integration into networks arises. The challenges of the future concern both the mobility of people and the transport of goods. A sustainable use of resources requires better organisation, through the interconnection of the different modes of transport, the integration of new technologies and the pooling of resources. This performance gain requires the design of agile and reconfigurable networks. The acceptance of such systems requires reliable, hazard-proof and reactive solutions. The LS2N, through its teams, offers key skills and experience to meet these challenges. Indeed, the quality of decisions depends first and foremost on the acquisition, exploitation and dynamic management of data evolution. Secondly, it involves modelling the behaviour and uses on the network. Tools and methods are needed to design an efficient network and evaluate its performance. It is necessary to calculate the flow routes and optimize the routes and schedules of the different modes of transport, while managing their synchronization. In this context, the integration of business rules, preferences or constraints that severely limit the space of solutions can be extremely complex. Tools and methods are also being developed to enable network management in a dynamic and uncertain environment.

 

The laboratory's strengths: the LS2N relies on its history of collaboration with car manufacturers, on the GIS ITS of which it is a member, on the national network created during the New Industrial France plan on autonomous vehicles, in consultation with the VEDECOM Institute, and locally on the ID4Car competitiveness cluster.



Copyright : LS2N 2017 - Legal notices - 
 -