This is the result of session 7 in my course on Social Robot Design (2025/2026). The content of this session was mainly about further designing a design tool to prototype the behavior of the robot, and also evaluate it on the robot itself. We
Session 7
Test the card game
Designing for robot expression through iterations of scenario-based interactions.
Option 1: Pick random cards
Expression/behaviour (puppeteering the robot)
We notice a big threshold to start actually acting.
It might help if we start by acting with the Miro (moving its parts).
- Round one: Pick a robot card and act it out with the Miro
- Round two: Pick the human card and act out the scenario
- Round three (optional but fun, repeat as much as you like): Pick a wild card to add a ‘saus’ to the scene
Option 2: Pick a student action card and let the players choose a robot action to match it
Scenario based. Acting is possible with Miro-E robot, but can be explored through group discussions.
- Round one: Pick one random human card
- Round two: Pick 5 random robot cards and lay them open so everyone can read them
- Round three: Talk about how the human will react to each of the robot cards. Rank them from best to worst. Talk about how the action on the worst robot cards could be improved to give a better outcome.
- Round four: Which card or combination of robot cards will result in the best outcome (could be all 5 cards, but a minimum of 1 robot card)?
- Round five: Try to get to a desired outcome with a minimum number of cards
- Round six (optional but fun, repeat as much as you like): Pick a wild card and see if it changes anything in the ranking you made, then you can repeat round five with the new ranking
Rejected option
Tried picking one robot card and 5 human cards, but that did not work as nicely because the human action always comes first in the timeline of the scene, so it’s more logical to choose how the robot reacts to the human’s action.
Feedback from the teacher
The interactions that are designed during the game need to be validated. So, after thinking of a scenario, something needs to happen to see if it will work out in this way in real life.
Currently it is too much like a generative suggestion box. It needs validation / evaluation of the scenario → need more worldbased/realistic ‘proof’, for example from puppeteering / acting / etc.
Idea: Create an acting and puppeteering evaluation manual(?)
Puppeteering / acting manual
Act out the scenarios that were created during the game.
Form to record the scenario playout
The scenario included the following cards:
| Human | Robot | Wildcard |
|---|---|---|
Rank the interaction for each of the criteria below from an observer’s perspective:
| 1 | — | 5 |
|---|---|---|
| Ineffective | Effective | |
| Demotivating | Motivational | |
| Confusing | Clear | |
| Unkind | Kind | |
| Awkward | Natural | |
| Oblivious | Empathic | |
| Unpredictable | Predictable | |
| Unreliable | Trustworthy |
Trying to make the Miro-e operational
Puppeteering it, Wizard of Oz style.
Design guidelines after getting the evaluation
After evaluating the game and the results, what steps can we take in our social robot design?
In the evaluation step, you encountered scenarios that are plausible/implausible and ranked them. Based on this ranking/evaluation, try to match those with the requirements of the designers (you!). Or, in other words, validate your requirements with the tool. A helpful tool is to create a table of requirements versus scenarios to systematically cross off requirements that are matched. You may also rank your requirements in order of importance; so when you encounter conflicting requirements, you know which requirement is ranked higher in importance and therefore should be met first.
For example, if one of the requirements is a small, portable robot, and in the evaluation step you have found out that you need a bigger robot (e.g. it has to move around), it might be good to revisit your list of requirements. Try to play around with the results you got from the game, and the physical/functional design of your robot.
This tool is meant as an iteration process to make sure time is not wasted when you are creating functional models before you have a good picture of what’s necessary. You can close the iterative process once you are at a physical/functional prototype stage by looking at whether your requirements are met when the user interacts with your robot. Document your changes when you change your design so you can backtrack what you have changed and why.
You are free to interpret the results as you see fit. There is no right or wrong.
This was done with:
- Liz van Ginderen (s27349745)
- Anna Hornman (s3056600)
- Oyindrila Sen Gupta (s3697762)
- Sarah Mans (s2306379)