Suscribirse por RSS
Banner ROA

pdf

Telepresence Reinforcement-learning Social Agent. (TERESA)

TERESA aims to develop a telepresence robot of unprecedented social intelligence, thereby helping to pave the way for the deployment of robots in settings such as homes, schools, and hospitals that require substantial human interaction. In telepresence systems, a human controller remotely interacts with people by guiding a remotely located robot, allowing the controller to be more physically present than with standard teleconferencing. We will develop a new telepresence system that frees the controller from low-level decisions regarding navigation and body pose in social settings. Instead, TERESA will have the social intelligence to perform these functions automatically.

The project's main result will be a new partially autonomous telepresence system with the capacity to make socially intelligent low-level decisions for the controller.

TERESA aims to develop a telepresence robot of unprecedented social intelligence, thereby helping to pave the way for the deployment of robots in settings such as homes, schools, and hospitals that require substantial human interaction.

Universidad Pablo de Olavide

The project aims to achieve substantially advance in the state of the art in socially intelligent cognitive robotic systems. A socially intelligent robot is one with the cognitive capacity to estimate human affective states and adhere to social norms when interacting with humans. the project focuses on an emerging class of robots called telepresence robots.

TERESA aims to meet four primary objectives:

  1. Detecting and understanding social behavior
  2. Socially normative telepresence robot behaviors
  3. Socially navigating the environment
  4. Learning social skills

The UPO research group will develop human-robot interaction models and navigation methods that exploit them,e.g. by developing path planning under uncertainties for efficient social navigation or by developing navigation policies for local online (re-)planning and execution.

  • Procedure handbook
  • Quality Assurance procedure
  • Report on user-requirements and detailed scenario for telepresence for the elderly
  • Web-site
  • Dissemination plan
  • Manually controlled telepresence system
  • Specification of data sets and open source software to be made available
  • Integrated feedback system
  • Iterative report on socially normative remote presence robot behaviors
  • First version of semiautonomous telepresence system
  • Final localisation component for multiple people detection
  • Models for social navigation planning
  • Library of macros and labeled data
  • Second version of semiautonomous telepresence system
  • Report on longitudinal effects of social participation through telepresence robots for elderly people
  • Elderly user-controller interface design
  • Demonstrator of controller's behaviour analyser
  • Demonstrator of behavior analyser for people in the robot's vicinity
  • Model for continuous dialogue monitoring
  • Third version of semiautonomous telepresence system
  • Demonstrator for learning pose policy
  • Global social path planning
  • Local navigation policies for execution and replanning using reinforcement learning
  • Report on final social intelligence performance evaluation
  • Report on the usage and usability of the telepresence robot for the elderly
  • Provision of vision data-set for use by the scientific community as well as open source software
  • Exploitation plan

The addition of socially intelligent features has the potential for a significant impact by making an entirely new set of venues accessible by mobile telepresence devices, venues where social intelligence adds significant value. Home care, education, and access to public places such as elderly day centers are examples of such venues. The home care market in Europe alone is well over 10 million homes.

 


Robotics, Vision and Control Research Group

Code PAIDI: TEP-151

Luis Merino. Investigador Principal. 

Universidad Pablo de Olavide

Budget of Andalusian group: € 341,115.00

http://aramis.upo.es/isa/lmercab/

  • Universiteit Van Amsterdam (Coordinator)
  • Universiteit Twente
  • Universidad Pablo de Olavide
  • Imperial College Of Science, Technology And Medicine
  • Giraff Technologies Ab
  • Centre Expert En Technologies Et Services Pour Le Maintien En Autonomie A Domicile Des Pesonnes agees
Keywords: Informatics Research, Telepresence robotics, Socially intelligent robotics, Robot navigation, Human-robot interaction, Computer vision, Machine learning, Reinforcement-learning
Duration: 36 months. December, 1th 2013 to November, 30th 2016
Project cost: € 3,889,900.00