In July of 2003, the first annual robotics summer school sponsored by the IEEE Robotics and Automation Society and the International Foundation of Robotics Research took place in Volterra, Italy. In this small picturesque town in the Tuscan province, approximately 40 faculty, students, and researchers gathered to share their knowledge of human-robot interaction in the school organized by professor Henrik Christensen from the Swedish Royal Institute of Technology (KTH) and professor Ronald Arkin from the Georgia Institute of Technology. During those five days, summer school participants heard lectures from six leading researchers in the field, discussed their own research, and worked hands-on with current research robots.
Thanks to the generous sponsorship provided by Science Applications International Corporation (SAIC), the European Robotics Research Network (EURON), and the Scuola Superiore Sant’Anna, the cost to attendees was low, allowing the school to garner participation from researchers around the world. Summer school participants from 15 countries in Europe, North America, Asia, and the Pacific Rim and were taught by professor Aude Billard from the Swiss Federal Institute of Technology in Lausanne, professor Cynthia Breazeal from the MIT Media Lab, professor Hiroshi Ishiguro from Osaka University and ATR, professor Erika Rogers from California Polytechnic State University, Dr. Jean Scholz of the United States National Institute of Standards and Technology, and Dr. Alan Schultz from the United States Naval Research Laboratory. Highlights from two of the lectures follow.
The School: HRI Evaluation
Dr. Jean Scholz highlighted an important area of current work in human-robot interaction evaluation in her lectures. She explained the history of human-computer interaction as it relates to HRI, expressed her views on the current state of evaluation in HRI, and laid out several metrics, methodologies, and examples of these in use. After several hours of detailed illustration of the principles of evaluation in HRI, Dr. Scholz had insured that anyone in the room without a background in evaluation had the necessary information to begin incorporating evaluation into their work. For those with more experience in HRI evaluation, the lecture presented many opportunities to reflect on the state of our field and consider how we can improve the types of analyses that are currently performed.
To insure that the methods described in the lectures were understood, Dr. Scholz led the participants through an hour-long evaluation example. Students watched video tapes from the 2003 RoboCup competition and recorded every "critical incident’’ they witnessedtimes when the robot could cause damage to itself, someone, or something else. After reviewing the video in small groups, Dr. Scholz led the group through a discussion of how this methodology can be applied to other interactions. Her lecture concluded with some concrete directions for the future of HRI and a call for the attendees to consider the importance of evaluation in their future research.
Professor Billard brought several of miniature humanoid robots, called Robota, that she and her students have been developing since 1997. These robots allowed students to try out several projects using the hardware and software tools provided by the Robota system. Dr. Billard had just lectured on imitation learning in humans as an inspiration for robot learning. Participants in the demonstration spent several hours linking the provided vision algorithms to the behavior system of the Robota dolls, defining the behaviors they wanted to happen based on the gestural input they made to the robot. In just a short time, these robotics researchers understood how to apply many of the theories presented in the lecture.
The concluding lecture of the week was a rapid-fire presentation by Dr. Hiroshi Ishiguro on android robots. Dr. Ishiguro explained his views of robots as an information medium. He expressed his belief that the best form for them to take is that of a human. He explained that when the appearance and behavior of a robot are close to that of a human, it is more natural for a person to interact with the robot. Ishiguro showed several examples of robots that he has built over the last several years at ATR and Osaka University.
During the lecture, Dr. Ishiguro presented his Robovie platform. His team has built numerous versions of the Robovie based on this basic architecture. The newest, Robovie-M, walked around during a break in the lecture, allowing students to see and control the robot first-hand. The lecture topics ranged from open problems in HRI, creating a soft skin to allow for more natural physical interaction to psychological evaluation of HRI, the combination of the social sciences and robotics, and the definition of the new field of android science. On this last topic, students learned that he has modeled two recent robots off of actual humans, attempting to mimic the appearance, responses, and behaviors of people in the quest for creating realistic android robots.
More than Work
In addition to the time spent listening to lectures and working on projects, participants took advantage of their surroundings. July in Tuscany is a beautiful time to be outside, which students and faculty found time for each evening, wandering the small, ancient Comune di Volterra. There were ancient buildings and gates, restaurants, and gelaterias to be explored every day of the week. The school’s schedule included tours of Volterra and the nearby Sienna, a city known for its medieval walls and large Palazzo Publico and the adjoining Piazza del Campo.
Over the course of five days at the end of July, over 30 researchers rapidly became acquainted with topics spanning the entire field of human-robot interaction. Just as important, they become acquainted with many of the leading researchers in the field, as well as the ambitious upcoming generation of HRI researchers. As any nascent field grows, these are the two of the most important factors that will help it mature into a solid domain. This HRI summer school clearly brought together a group of people excited about the work they are undertaking and who are able and willing to work together to help HRI coalesce into a stable and growing community of researchers. After leaving Volterra, there are already plans for a follow-up workshop in the summer of 2005 to once again bring together young researchers in HRI.
Cory D. Kidd is a Ph.D. student at the MIT Media Lab. His work in human-robot interaction is focused on finding solutions to current healthcare issues using sociable robots. firstname.lastname@example.org
©2005 ACM 1072-5220/05/0300 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2005 ACM, Inc.