Robotics Research Lab
CRES
USC Computer Science
USC Engineering
USC
/ Research / Projects / Socially Assistive Human-Machine Interaction for Improved Compliance and Health Outcomes

Overview Project Details Publications Support People

Overview

This project is focused on developing and evaluating human-machine interaction techniques, especially socially assistive robotics (SAR) techniques, that influence the user to engage in wellness-promoting behaviors involving physical, cognitive, and social activity. Our work for this project focuses on two types of wellness-promoting human-machine interactions: 1) exercise sessions and 2) socializing sessions. In the former, the SAR provides exercise monitoring, coaching, and motivation, while in the latter, the SAR provides social contact and friendly reminders and encouragement. The two types of interaction share the common underlying goal of influencing behavior. We take a twofold approach to reaching that goal: 1) we use steering to learn how to affect the user’s behavior in the short term during an interaction; and 2) we use motivation to affect the user’s longer-term behavior over the course of an interaction and help to maintain their engagement in the interaction. This work has developed and validated socially assistive techniques for influencing behavior by steering interaction dynamics and by encouraging performance. The embodiment of the SAR is leveraged to maximize engagement and compliance; therefore the work for this project also focuses on natural, persuasive and engaging embodied communication.

Project Details

Robot Motivator: Towards Adaptive Health Games for Productive Long-Term Interaction

This research focuses on the development, evaluation, and user testing of a Socially Assistive Robot (SAR) exercise coach designed to motivate and engage older adults in a seated aerobic exercise task. Our SAR system approach incorporates insights from psychology research into intrinsic motivation and contributes clear design principles for SAR-based therapeutic interventions. Through our user studies, we have examined several different aspects of wellness-promoting socially assistive systems, including a comparison of an embodied versus a non-embodied (computer-based) system, the role of praise and relational discourse, and the effect of varying user autonomy.

Human Activity Monitoring for Socially Assistive Interaction with Older Adults

In this project, we extended our previous work with the elderly performing "chair exercises" guided by a socially assistive robot. The exercise scenario utilized a socially assistive robot to instruct, evaluate, and encourage users to perform simple arm gesture exercises. The scenario was one-on-one, allowing the robot to focus its attention on the single user in order to provide timely, accurate feedback, and to maximize the effectiveness of the exercise session for the user. In the set up, the user was seated in a chair in front of the robot; the user and robot faced each other. The developed probabilistic activity monitoring systems affords the robot the ability to track the user’s arm movements; the use of the Kinect sensor (as opposed to a monocular camera) is an extension of our previous work, in that the robot and arm motion is no longer restricted to the sides of the body (i.e., non-planar)

Automated Proxemic Behavior Recognition and Production

This research investigates proxemics in human-robot interaction (HRI). Proxemics is the study of the dynamic process by which people position themselves in face-to-face social encounters. This process is governed by sociocultural norms that, in effect, determine the overall sensory experience of each interacting participant. To facilitate situated and mobile HRI, this research seeks to develop functional and socially appropriate probabilistic computational models of proxemics for the purposes of both autonomous proxemic behavior recognition (of one or many people) and autonomous proxemic behavior production (by a sociable robot).

Active NAO! Combating Childhood Obesity with Robot Companions

In this project we are developing methods for creating a cohesive character for a Socially Assistive Robot (SAR) exercise buddy with the goal of motivating increased exercise effort. We are using a SAR in a teammate role on a peer level interaction with a user engaged in physical exercise. We are currently developing the SAR exercise buddy system for circuit training with obese and overweight youth age 11-14.

Spatial Language-Based Human-Robot Interaction

This work presents a novel methodology that allows service robots to interpret and follow spatial language instructions, with and without user-specified natural language constraints and/or unvoiced pragmatic constraints. This work contributes a general computational framework for the representation of dynamic spatial relations, with both local and global properties. The methodology also contributes a probabilistic approach in the inference of instruction semantics; a general approach for interpreting object pick-and-place tasks; and a novel probabilistic algorithm for the automatic extraction of contextually and semantically valid instruction sequences from unconstrained spatial language discourse, including those containing anaphoric reference expressions.

Publications
Search for a Publication

On-Line Publications

Conference Papers
Aaron B. St. Clair and Maja J. Matarić. "Coordinating Communication in Human-Robot Task Collaborations". Refereed Workshop Workshop on Learning Plans with Context from Human Signals at Robotics: Science and Systems , Berkeley, CA, USA, Jul 2014. (.pdf)(Details)
Ross Mead and Maja J. Matarić. "Perceptual Models of Human-Robot Proxemics". To appear in 14th International Symposium on Experimental Robotics, 2014 (ISER 2014) (ISER), Marrakech/Essaouira, Morocco, Jun 2014. (Details)
Juan Fasola and Maja J. Matarić. "Interpreting Instruction Sequences in Spatial Language Discourse with Pragmatics towards Natural Human-Robot Interaction". To appear in International Conference on Robotics and Automation (ICRA), May 2014. (.pdf)(Details)
Aaron B. St. Clair and Maja J. Matarić. "Studying Verbal Feedback in Human Collaborations to Inform Robot Speech Production". To appear in 2014 International Conference on Collaboration Technologies and Systems - 5th International Workshop on Collaborative Robots and Human Robot Interaction (CTS 2014), Minneapolis, MN, USA, May 2014. (.pdf)(Details)
Ross Mead and Maja J. Matarić. "Probabilistic Models of Proxemics for Spatially Situated Communication in HRI". In 9th ACM/IEEE International Conference on Human-Robot Interaction (HRI 2014) Algorithmic Human-Robot Interaction Workshop (AHRI), Bielefeld, Germany, Mar 2014. (.pdf)(Details)

Journal Papers
Juan Fasola and Maja J. Matarić. "A Socially Assistive Robot Exercise Coach for the Elderly". In Journal of Human-Robot Interaction, 2(2):3-32, Jun 2013. (.pdf)(Details)
Ross Mead, Amin Atrash and Maja J. Matarić. "Automated Proxemic Feature Extraction and Behavior Recognition: Applications in Human-Robot Interaction". In International Journal of Social Robotics, 5(3):367-378, 2013. (.pdf)(Details)
Conference Papers
Juan Fasola and Maja J. Matarić. "Using Semantic Fields to Model Dynamic Spatial Relations in a Robot Architecture for Natural Language Instruction of Service Robots". In IEEE/RSJ International Conference on Intelligent Robots and Systems 2013 (IROS 2013), Tokyo, Japan, Nov 2013. (.pdf)(Details)
Juan Fasola and Maja J. Matarić. "Using Spatial Semantic and Pragmatic Fields to Interpret Natural Language Pick-and-Place Instructions for a Mobile Service Robot". In International Conference on Social Robotics (ICSR 2013) , Bristol, UK, Oct 2013. (.pdf)(Details)
Juan Fasola and Maja J. Matarić. "Modeling Dynamic Spatial Relations with Global Properties for Natural Language-Based Human-Robot Interaction". In 22nd IEEE International Symposium on Robot and Human Interactive Communication (RO-MAN), Gyeongju, Korea, Aug 2013. (.pdf)(Details)
Ross Mead, Stefan Zeltner and Maja J. Matarić. "Integrating ROS into Educational Robotics: Bridging the Gap between Grade School and Grad School". In Global Conference on Educational Robotics (GCER), 2013 Norman, Oklahoma, Jul 2013. (.pdf)(Details)
Aaron B. St. Clair and Maja J. Matarić. "Role-Based Coordinating Communication for Effective Human-Robot Task Collaborations". In 4th International Workshop on Collaborative Robots and Human Robot Interaction (CR-HRI 2013), May 2013. (.pdf)(Details)

Journal Papers
Juan Fasola and Maja J. Matarić. "Using Socially Assistive Human-Robot Interaction to Motivate Physical Exercise for Older Adults". In Proceedings of the IEEE - Special Issue on Quality of Life Technology, 100(8):2512-2526, Aug 2012. (.pdf)(Details)
Conference Papers
Juan Fasola and Maja J. Matarić. "Using Spatial Language to Guide and Instruct Robots in Household Environments". Refereed Workshop AAAI AAAI Fall Symposium: Robots Learning Interactively from Human Teachers (AAAI 2012), Arlington, VA, Nov 2012. (.pdf)(Details)
Ross Mead. "Space, Speech, and Gesture in Human-Robot Interaction". Doctoral consortium, International Conference on Multimodal Interaction (ICMI), 2012 333-336, Santa Monica, California, Oct 2012. (.pdf)(Details)
Juan Fasola and Maja J. Matarić. "Socially Assistive Robot Exercise Coach: Motivating Older Adults to Engage in Physical Exercise". In 13th International Symposium on Experimental Robotics, 2012 (ISER 2012), 2012. (.pdf)(Details)
Technical Reports
Ross Mead, Amin Atrash and Maja J. Matarić. "Representations of Proxemic Behavior for Human-Machine Interaction". NordiCHI, Copenhagen, Denmark, Workshop, Oct 2012. (.pdf)(Details)

Journal Papers
Ross Mead, Amin Atrash and Maja J. Matarić. "Automated Analysis of Proxemic Behavior: Leveraging Metrics from the Social Sciences". Refereed Workshop Robotics: Science and Systems, 2011, Jul 2011. (.pdf)(Details)

Support

This work is supported by National Science Foundation (NSF) grant titled "Socially Assistive Human-Machine Interaction for Improved Compliance and Health Outcomes" - Award number: IIS-1117279.

Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

People

Maja Mataric (PI)
Amin Atrash
Ross Mead
Elaine Short
Katelyn Swift-Spong

Past Contributors

Juan Fasola