[robotics-worldwide] [meetings] Call for Papers for ARSO 2015 Special Session on Human-Robot Interaction for Autonomous Cars

Dirk Wollherr dw at tum.de
Mon Mar 9 05:24:06 PDT 2015


Call for Papers:  Special Session on Human-Robot Interaction for Autonomous Cars
At the 2015 IEEE International Workshop on Advanced Robotics and its Social Impacts (ARSO 2015) http://arso2015.inria.fr <http://arso2015.inria.fr/>
1-3 July 2015, Lyon, France


Important Dates:

31 March 2015: Paper submission 
4 May 2015: Acceptance notification
30 May 2015: Final paper submission


Motivation and Objectives:

The issue of human-robot interaction for autonomous vehicles is two-fold: on the one hand side, autonomous vehicles must interact with people inside the car, i.e. the driver or passengers, and on the other hand with people on the street. This organized session aims at highlighting issues and solutions in human-automobile interaction with respect to both aspects.

While some manufacturers envision fully automated vehicles with no possibility for passengers to influence the driving behavior of the car, others plan to allow handover between automated driving, human driving, and various degrees of driver assistance. In all of these situations, there the vehicle must communicate with people and convey its intention, as well as understand the intention of humans in order to react to them. In the case of passive passengers, i.e.\ fully autonomous driving, the driving action must impart the feeling of safety and predictability of the action. On the other hand, even in fully autonomous driving, it would be desirable, if the vehicle would adapt the driving style to the passenger needs, e.g. slow driving for sightseeing vs. goal directed driving in order to reach a given destination at a desired time. Such adaptation of the driving style on the other hand requires understanding of the passenger intention and thus some sort of communication.

In cases of shared control over the vehicle - especially with hand-over of control between human and car - it must become obvious to both actors, who expects which actions by the other. Misunderstandings, wrong interpretation of signals or even delayed reaction to action requests can result in dangerous situations or even accidents. Therefore, an intuitive and easy-to-understand, but explicit and unambiguous interface to the human is desirable. Furthermore, the interface should be adaptive and able to correct errors. 

A second issue in autonomous driving is to convey the intention or goal of an autonomous car to other (human) drivers or road users, such as pedestrians, cyclists, etc. While autonomous vehicles can easily compute collision-free trajectories, assess the collision probability and thus navigate safely even in crowded urban environments, the motion must still be understandable and predictable to humans in order to avoid confusion or panic reactions.

The proposed organized session will gather the state-of-the-art in human-vehicle interaction - both, passengers in the car and road-users outside of the car - and present solutions to these interaction issues.


Confirmed Speakers:

Philipp Kerschbaum: “Designing the user interface for highly automated cars - concepts, studies and challenges”
Annemarie Turnwald: “Similarity assessment fo human trajectories ofr enhanced readability in human-robot interaction tasks"


Organizer:
Dirk Wollherr, Technische Universität München

-- 
----------------------------------------------------------------------
Prof. Dr.-Ing. habil. Dirk Wollherr
Lehrstuhl für Steuerungs- und Regelungstechnik/
  Chair of Automatic Control Engineering (LSR)
Technische Universität München
D-80290 München, Germany
Tel.:  +49 89 289-23401, Fax: +49 89 289-28340
mailto:dw at tum.de <mailto:dw at tum.de>
WWW:   http://lsr.ei.tum.de <http://lsr.ei.tum.de/>


More information about the robotics-worldwide mailing list