Breakout group C1

1
Breakout group C1

• Monica Nicolescu
• Maja Mataric
• Gunter Niemeyer
• Benjamin Kuipers
• Warren Dixon
• Jill Drury
• Jana Koseka
• Brian Scassellati

2
What must researchers and NSF do to achieve
measurable results in HRI?

• Established NSF program, long term start with
  short-term funding
• For interdisciplinary research, it takes time to
  get tangible results
• Use measures
• HR team does better than H and R alone (improve
  human performance)
• User studies get new design criteria
• Get to more visible (public, external) situations

3
What must researchers and NSF do to achieve
measurable results in HRI?

• Get a better understanding of the application
  areas, gain insight into different domain
• To what extent should we have an integrated
  system
• Prove the hypothesis of the work
• Performance measuring is subtle
• what measures are needed?

4
What must researchers and NSF do to achieve
measurable results in HRI?

• Set some challenges (common platform, set goals,
  set interfaces, generic model of the robot)
• Standard platform
• Wheelchair, humanoid
• Are we at the point where we can make these
  challenges? (yes, for some domains wheelchair)
• Dont want to fall into one category

5
What must researchers and NSF do to achieve
measurable results in HRI?

• What do we need to do to meet the needs of social
  interaction
• Should we work on so many diverse things
• Possibly better to work on individual parts
• There are common issues across different
  platforms
• Design systems with more sophisticated
  interaction
• Better interfaces
• Better use of technologies that are available
• What is the timeline for deploying robots to
  real-world applications

6
What must researchers and NSF do to achieve
measurable results in HRI?

• Does momentum means moving together -gt common
  applications, theories
• Continue collaborations just like we did so far
• Clusters are emerging focal points of research
• Goals (Focal points)
• Manipulation
• Mobility (structured/unstructured domain)
• Social interactions

7
What must researchers and NSF do to achieve
measurable results in HRI?

• Types of interaction
• Dexterous
• Social
• Navigation
• Roles
• R as tool/peer/teacher
• Methods (cross-cutting)
• Natural language
• Perception
• Representations
• Properties of instrumentations (what can be done
  with a particular interface)

8
Name a few major applications that will dominate
the use of HRI in research results

• Follows from previous question
• Make some structure in the field (above
  dimensions)

9
How do these applications demonstrate the use of
HRI research results?

• H R perform better than alone
• Comfort, energy use, speed, other metrics
• Need to define the applications
• Need another dimension to the above spectrum
• elder care, therapy, search and rescue

10
What needs to be done to establish HRI as a field
and build a strong HRI community

• HRI conferences, HRI journals, targeted funding
• Industry involvement
• HRI textbook/handbook