Back to cases
DELIVERED2019Intelligent hardware · HRI

Service robots: trying the interaction before building the feature

From waiting for a finished feature to testing understanding now: separate interaction stages and simulate key responses before further development.

My role · Senior user research, experimental and field-study design, and method standardization

Get the method templates

Reframing the problem

If we only ask: How can a robot interaction feel more natural?

The more useful question: At each stage, do people know what they can do and recognize the response? Separate understanding the interaction from implementing it, then simulate key moments.

“Make it natural” was not enough to build from

A service robot communicates its state through sound, motion, and expression. Words such as natural or friendly do not tell an engineer what to change.

As a senior user researcher in 2019, I contributed to experiments and field studies of these cues, their effects on understanding and interaction, and the team’s research methods.

Breaking down the encounter

I separated interaction into attraction, initialization, purposeful interaction, and ending. Each stage has a concrete question: did someone notice it, know how to begin, repeat a command, or realize the encounter was over?

Sound and motion could then be discussed in terms of their job at that moment. Completion rates alone miss encounters that never started.

Testing with what was available

For unfinished functions I used Wizard-of-Oz: a researcher simulated responses behind the scenes so participants could try the proposed interaction. We could examine understanding and acceptance before committing to implementation.

This tests an interaction assumption. It does not establish that automated recognition or response technology works. Appropriate research authorization and explanation are needed, and controlled conditions and field use require separate observation.

My contribution and result

I contributed to study design, fieldwork, and documentation of interaction stages, variables, exceptions, and recording methods. Specific feature decisions, samples, and numerical results are private.

It is an example of an approach I often use: find an available way to test the key assumption while the complete product is still out of reach.

PROTOTYPE / TWO LAYERS
Experience before implementationThe frontstage preserves the start, task, and end of the user experience. A researcher simulates unfinished responses backstage. Test understanding before choosing implementation.FRONTSTAGE / EXPERIENCEBACKSTAGE / SIMULATIONSTART · TASK · ENDWizard-of-Oz
  1. Frontstage · ExperienceStart, perform the task, end
  2. Backstage · SimulationA person follows response rules
  3. Observe & decideTest understanding before implementation

Observe understanding · Simulate the responseConceptual model

Materials to use in your own work

These blank templates and instruction examples were prepared for this website. Download and adapt them to your task. They contain no client data and are not original project deliverables or evidence of results.

Notes / Thinking