|
|
|
FJA@ROMAN2016 > Program
9:00-9:15 Introduction (Aurélie Clodic, LAAS-CNRS) 9:15-9:30 On the Communicative Aspect of Human-Robot Joint Action (Ross A. Knepper, Cornell University) 9:30-9:45 What common ground do we need (most) for speaking to a robot?(Kerstin Fischer, University of Southern Denmark) 9:45-10:00 Interdependence as a Framework forCommon Ground in Joint activity(Matthew Johnson, IHMC) 10:00-10:15 New Challenges for Human-Robot Collaboration in an Industrial Context: Acceptability and Natural Collaboration(Eva Coupete, Vincent Weistroffer and Olivier Hugues, Mines ParisTech and CEA-LIST)
10:30-11:00AM Coffee Break ------------------------------------------------------------------------------------------------------------------------ 11:00-12:30PM---------------------------------------------------------------------------------------------------------------------------------------- 11:00-12:00 Speakers and their authority (Herbert H. Clark, Stanford University) When people speak, they make commitments based on various types of authority. In the simplest exchanges, people act on their own authority. At the door, the hostess says “Welcome” and the guest responds “Thank you.” In other exchanges, one or both parties act on derivative authorities. At Macy’s, a customer asks “How much is that?” and the clerk replies “Twenty dollars.” The customer acts on her own authority, but the clerk acts partly on an authority granted by Macy’s. But what if the speaker is a virtual agent or robot? In saying “Welcome,” the hostess expresses her own pleasure in seeing the guest. But when the PA system at the Denver airport announces “Welcome to Denver,” whose pleasure is being expressed and on whose authority? In joint actions, people attribute various types of authority to their coparticipants. But when a co-participant is a virtual agent or robot, how do they do that, and to what end?
13:30-15:00PM-------------------------------------------------------------------------------------------------------------------------------------------- 13:30-15:00 Interactive session We plan to organize two panel discussions. The first one dedicated to common ground needed around joint action execution, its components, its installation and the way to model it in a human-robot context. The second one dedicated to the evolution of this common ground along the joint action execution and particularly coordination and alignment issues that need to be taken into account. 15:00-15:30PM Coffee Break------------------------------------------------------------------------------------------------------------------------- 15:30-17:00PM------------------------------------------------------------------------------------------------------------------------------------------- 15:30-16:30 Common ground in coordinated action (Tamara Lorenz, University of Cincinnati) If we want a person to be engaged in a natural interaction with a robot, it is important that the robot’s behavior matches the human’s expectations. Just like forming common ground in dialogue, a common ground has to be formed in order to enable safe and intuitive interaction. Approaching this problem in human-human interaction, it was found that people tend to synchronize their movements to distinct phase relations that allows not only for efficient individual action, but also for efficient coordination. If this should also be enabled in human-robot interaction, an interaction model is required to enable the robot to feed into this coordination process successfully. We proposed an interaction model based on coupled Kuramoto-oscillators and could demonstrate that such a model does not only allow for coordinated action, but can also be used to test the establishment of behavioral common ground, which can be measured by the extent to which the human engages in this interaction. 16:30-17:00 General Panel |
| Online user: 60 |
|
