Operational transparency
What the robot does, what it can’t do, and what it does with what it sees and hears.
Summer internship, ACM OzCHI 2025
Human robot interaction
A Pepper robot in the corridors of the University of Oulu, and a study of what students and staff would need before they trusted one on campus.
01 The question
A robot on campus is only useful if people trust it enough to use it. The study asked the people who would be using it what that trust depends on.
02 The study
- participants
- 59
- questionnaires
- 44
- interviews
- 17
- minutes per interview
- 60
-
01
Questionnaire 1
Attitudes towards social robots, and what people would use one for.
-
02
Questionnaire 2
The same ideas again, now with the trust concerns behind them.
-
03
Interviews
An hour each, on what trust depends on in this particular place.
Pepper was there for both the questionnaires and the interviews. It followed people with its eyes, moved its hands while it spoke, offered five topics on its chest tablet, answered by voice and reacted when its head or hands were touched.
03 What trust depends on
Knowing what it does, and who is behind it
Operator transparency
Who is behind it, who answers for it, and who can see the data it collects.
-
“It is difficult to know who is behind her”
P22 -
“It all depends on the programming in the background”
P43
The paper adds these to Hancock et al.’s factors of trust in robots, a model built for robots working in teams, and proposes transparency and security as an eighth characteristic in Baraka et al.’s framework for describing social robots.
04 What people wanted
61 ideas, seven kinds of help
-
01
Navigation
The most common wish. The campus is large and its buildings look alike.
-
02
On-campus assistant
Answers around the clock when the info desk is closed.
-
03
Social assistant
Finding study partners, classmates and staff; translating on the spot.
-
04
Promotion
“Hey, Pepper, what can I do this afternoon at the university?”
-
05
Safety and security
Spotting emergencies and following a protocol to raise the alarm.
-
06
Wellbeing
Someone to talk to, relaxation techniques, stretching breaks.
-
07
Entertainment
A photographer, a mascot, the occasional high five.
05 Four questions
What any campus robot will be asked
-
Does it save me time?
If a phone is quicker, the robot loses. It has to be near, fast and fluent.
-
Is it obtrusive?
No noise, no unwanted approaches, no feeling of being watched. There when needed, invisible otherwise.
-
Is it fun?
Not required, but welcome, as long as the fun never gets in the way.
-
How does it relate to the world around it?
A robot replaces or extends something already there, from notice screens to sign language.
06 My part
I joined the group as a summer intern during my master’s. I helped develop the robot, above all its text-to-speech and its movements, so that it came across as friendlier and more human, and I helped gather the responses. The study was the group’s; I am one of the paper’s eight authors.
What I contributed
Robot features
Helped develop Pepper’s features for the study, above all its text-to-speech, so it could speak to the people taking part.
Friendlier movement
Gestures and body movement that made the robot feel more human and easier to approach.
Gathering responses
Helped collect the responses the research was built on.
The paper
One of eight authors on the ACM OzCHI 2025 paper. Read it (opens in a new tab)
- Pepper robot
- Text-to-speech
- Robot movement
- Response collection
07 The paper
Factors of Trust for Successfully Adopting Social Robots on the Campus
Read the paper ↗ (opens in a new tab)
Next project
NeuroClima ↗