Experiments with Social Robots in the Wild.
Summary of the 2nd chapter of the Book by Takayuki Kanda & Hiroshi Ishiguro “Human-Robot Interaction for Social Robotics.
5. ScienceMuseum-Motivation
Robovies interacted with the visitors and showed them around to exhibits
according to information from ubiquitous sensors.
During the two month experiment, visitors enjoyed interacting with the robots and
highly appreciated them.
1. for visitors opportunity to experience the advanced technologies:
fundamental purpose of a SM - easily deploy research to a real environment.
2. targeting people interested in robots: SM one of the best choices for
collecting feedback and examining the HRI.
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
WHAT?
WHY?
6. ScienceMuseum-ExperimentalSettings
sensors to work in a complex, crowded environment:
cameras + wireless tag readers (RFID)
coordinates for the robots & call visitors by names, recommendations based on movement history
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
● 20 RFID tag readers (2 on robots), 3IR sensors, 4 video cameras.
● 1 humanoid - guide to the exhibits, 2 stationary - explained the exhibits, 1 robot greeting and
asking them to return their RFID tags
● central database - interaction data between robots and visitors: Ethernet
7. ScienceMuseum-SENSORS
For friendly interaction with visitors, robots need information about them
(interest increases when the machines call them by name) ->
SENSORS:
1. robots can act more intelligently.
2. robots contain an RFID tag readers -> not only interactive robots but also part of the sensor
system.
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
10. ScienceMuseum-Experiment
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
2-month period, visitors 91107, with RFID tags 11927 and 2891 returned questionnaires
Questionnaire by ranking five factors on a scale of 1-to-5
Moreover, visitors asked to:
give opinions on robots
- We had a really good time. - I had fun because the robots called me by name. - We felt close to the robots.
describe their favorite robot behavior
such as hugging, the calling out of names, and so on (basic elements of human society).
11. ScienceMuseum-Conclusion
● human-like body attracts people and enables human-like interaction
(shaking hands, pointing...)
● tactile interaction capability - effective
● lack of speech recognition
● human identification (this case-RFID) must be robust (visual and
auditory difficult in noisy real world!)
● infrared camera - exact position of the robot - very helpful in crowded
environment
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
12. citation:
Humanoid Robots as a Passive-Social Medium -
A Field Experiment at a Train Station
by K. Hayashi, D. Sakamoto, T. Kanda et.al.
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
FieldTestataTrainSTation
basedonthebook“HRIforSocialRobotics”
13. TrainSTation-Motivation
Using robots as a passive social medium, in which multiple robots
converse with each other
Field experiment at a train station for eight days to investigate the effects
of a passive social medium
1. If the robot(s) is “interactive” with people by bowing to them before
announcing the information, the people will get a stronger feeling of
being addressed by the robots
2. If a robot is passive toward people without reacting to them, people
will pay more attention to the information coming from the robot(s)
3. People are more likely to stop to listen to the robot’s conversation in
a two-robot condition than in a one-robot condition
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
WHAT?
WHY?
15. The robot(s) announces various features of the new
train line
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
TrainStation-Scenario
17. TrainSTation-Results
5900 people were observed
Questionnaire (163)
Analysis of behaviors - ignoring (370), noticing (260),
stopping to watch (110), staying (1), touching (1),
changing course to investigate (50), talking about robot
(s) (7), watching with child (15)
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
18. TrainSTation-Conclusion
social expression vs. limited-realistic interactivity
passive-social medium - most effective way of
attracting people’s interest in the information
interactivity - useful in giving people the feeling of
talking with the robots
FieldTests-ObservingPeople´sreaction basedonthebook“HRIforSocialRobotics”
19. Passive vs. interactive
social robots, which one
is better and why?
InspirationalQuestions:
let´sdiscuss
Why to go outside of
laboratories to perform
experiments?
What can the sensory
network (smart/intelligence
space) serve for?
How do we measure the
quality of the human-
robot interaction?
What are the (dis)
advantages of using
autonomous versus
teleoperated robots in
the first real-world HRI
experiments?
How to design an
experiment in the wild?
What is the first thing to
do when designing an
experiment?