Humanoid Robots as Helpful Partners for People

Organized by Cynthia Breazeal, MIT Media Lab

cynthiab@media.mit.edu

Overview

The aim of this workshop is to discuss the state-of-the-art in social human-humanoid interaction where the robot an human engage one another as peers. Throughout the workshop, we will focus on the core scientific questions as well as key engineering issues related to the design, implementation and use of such robots for a broad range of applications ranging from health care, entertainment, to planetary exploration, and beyond. Leading researchers from academic, commercial and government organizations will present their experiences and lessons learned from real-world systems. The workshop will culminate in a panel discussion to identify key scientific and technical challenges, likely future advances and potential fields of application.

This workshop is designed to benefit researchers interested in developing effective social humanoid systems, scientists who want to understand the state-of-the-art in social robots, and professionals planning collaborative human-humanoid applications.

The workshop wil take place on November 10, from 2pm-6pm. Presentation schedule listed below.

Description

Many of the earliest motivations for developing humanoid robots centered on creating robots that operate within human spaces and play a beneficial role in the daily lives of ordinary people. Robots, however, have traditionally been viewed as a tool: a device capable of performing tasks on (human) command. In particular, teleoperation of dexterous humanoid robots has proven to be cognitively demanding and requires a highly trained expert. Though this "robot as tool" approach suffices for some domains where the human is a specialist, it is sub-optimal for tasks that require significant human-robot teamwork --- or for scenarios where the user is a typical consumer. Such applications include collaborative exploration, construction, care of the elderly, providing assistance for the physically disabled, entertainment, domestic help, education, and more.

To support these applications, there has recently been a significant effort to develop humanoid robots that function more "naturally." Such robots support human social cues and are designed to interact with a wide range of people of different ages, gender, education, etc. Rather than being operated as mere tools, these "social robots" are designed to engage people as partners, if not peers. They interact socially with people and may express emotions, may communicate with natural language and gesture, and may exhibit a distinctive personality.

The purpose of this full-day workshop is to provide an international forum to examine and explore the state-of-the-art in social humanoid robots. In this workshop, we will discuss the design of and study the use of socially interactive humanoids in a broad range of applications. The objectives for this workshop are to:

  • foster technical discussion among leading researchers

  • discuss design elements required for social humanoid robots including appearance, social cognition, human modeling, communication, interaction, and social learning

  • review common problems, challenges, and potential solutions

  • identify core scientific questions pertaining to the realization of socially intelligent humanoids

  • identify directions for future research and potential applications

  • educate the audience on the difficulties inherent in social humanoids, especially designing to support human social intelligence and interaction

Following the technical presentations, a panel session will be held to address the following questions:

  • What types of affordances and feedback are crucial for transparent communication between humans and humanoids?

  • What capabilities are necessary to support human-robot teamwork?

  • How to design robots that learn quickly from natural human instruction for acquiring new tasks and skills while "on the job"?

  • What are appropriate models for human-robot interaction? Is it useful to mimic or emulate human-human interaction? Are there other models that might also provide for effective personal interaction?

  • How can we evaluate social robots? Which metrics should be used? How can we measure social "effectiveness"?

  • What are the underlying social, philosophical, and ethical issues that may influence future technical development?

  • How to design humanoids that can establish and maintain a beneficial and long term relationship with people?

Schedule

2:00-2:20pm Cynthia Breazeal, MIT Media Lab

cynthiab@media.mit.edu

Welcome & Opening Comments

Beneficial Relationships with Robots

2:20-2:40pm Fred Rhenmark, NASA Johnson Space Center

fredrik.rehnmark1@jsc.nasa.gov

Evolving the NASA/DARPA Robonaut into an EVA Team Member

2:40-3:00pm Sherry Turkle, MIT STS Program

sturkle@media.mit.edu

Robot as Rorschach: Children and the Elderly

3:00-3:20pm Hideki Kozima, National Institute of Information and Communications Technology

xkozima@nict.go.jp

Humanoid and non-humanoid robots in company with children

3:20-3:40 Discussion

Roles of Robots in Human Society and our Relationship with Them

3:40-4:00pm Coffee Break.

4:00-4:20pm Maja Mataric, University of Southern California

mataric@pollux.usc.edu

Towards Articulated and Expressive Assistive Robots

4:20-4:40pm Michael T. Rosenstein, Andrew H. Fagg, and Roderic A. Grupen, UMASS Amherst

mtr@cs.umass.edu

Teleoperation as a Tool For Developing Humanoid Partners

4:40-5:00pm Atsuo Takanishi, Waseda University

takanisi@waseda.jp

Humanoid Robot Research at Waseda University: Towards Harmonized Human-Robot Society in Japan

5:00-5:30pm Candy Sidner & Christopher Lee, Mitsubishi Electric Research Laboratories

sidner@merl.com, lee@merl.com

Conversations with Robots

5:30-5:55pm Discussion

Key Scientific Questions and Technical Challenges for Social Robots

5:55-6pm Closing Comments & Adjourn for Reception