Zum Hauptinhalt springen

Presence by Design

For children facing cancer or serious illness, missing school means more than falling behind — it means losing friendships, routines, and the sense of normality that matters most during treatment.

Telepresence robots promise to bridge this gap. But technology alone isn't enough. If a child can't control what their classmates see, what their teacher hears, or what enters their private space — participation becomes exposure.

PRIVATAR is a research project that puts privacy at the center of this challenge. Published in the CERN IdeaSquare Journal of Experimental Innovation, this paper by the chilli mind team presents a UX framework in which data protection isn't a legal checkbox — it's the foundation of trust, agency, and meaningful inclusion.

Through child-centered interface design, spatial privacy boundaries, role-based controls, and AI-generated digital user twins, PRIVATAR demonstrates how informational self-determination can be designed — visually, intuitively, and at the right cognitive level for young and vulnerable users.

Privacy as empowerment. Not as a barrier.

Privacy must be visual

Abstract consent forms and dense legal language don't work for children — and in high-stress situations, they don't work well for adults either. The PRIVATAR interface replaces text-heavy privacy notices with a consistent visual language: purpose-designed icons, clear status indicators, and the "Sendbox" — a live display that shows exactly which data is currently being transmitted from home or hospital into the classroom. Audio active. Camera on. Avatar visible. Every state is immediately readable at a glance, giving children genuine awareness and control without requiring them to navigate settings menus or parse abstract concepts.

Spatial privacy through no-go and no-view zones

Privacy isn't only about data flows — it's also about physical space. In a classroom, certain areas carry implicit sensitivity: a teacher's desk during a private conversation, a corner used for small-group work, spaces where other children might be captured without consent. PRIVATAR encodes these boundaries directly into the robot's behavior. No-go zones stop the robot automatically at defined perimeters. No-view zones allow movement but restrict or blur visual transmission. When a boundary is reached, the child receives immediate, child-friendly feedback explaining what happened and what options are available — turning an abstract data protection rule into a tangible, understandable moment of control.

Core findings:

How can UX for privacy act as an enabler of participation?

When children can control what is visible, audible, and present, privacy stops being a barrier and becomes a foundation for trust and psychological safety. Self-determination is what makes showing up possible.

Which interface patterns operationalize privacy-by-design?

Three patterns emerged: the Sendbox provides a live view of all active data transmissions. No-go zones stop the robot automatically at defined boundaries; no-view zones allow movement but blur visual output. A tested set of pictograms ensures key states — camera, microphone, avatar — are instantly readable by children.

How do AI-generated digital user twins support the design process?

They enable exploration of rare, ethically sensitive scenarios — third parties appearing on camera, conflicting stakeholder preferences, cascading in-lesson events — without involving vulnerable users directly. Real user involvement was reduced by around 40%.

Four roles. One system. No one left unprotected.

Privacy in a telepresence system isn't a single switch — it's a layered set of decisions made by different people at different moments. A parent configuring defaults before the school day begins. A teacher adjusting volume during group work. A child choosing whether to raise their hand or step back from the camera. The PRIVATAR application structures these decisions into four clearly defined use cases — each mapped to the right person, at the right moment, with the right level of control.

UC 1 — Configuration & Setup

Handled by parents or a personal supporter. Fundamental default settings are defined here: which data is sent to the classroom, and which features — camera, microphone, face recognition, localization — are active.

UC 2 — Participation in Lessons

The child's primary interface. Controls presence, visibility, and audibility in real time. Also covers hand raising, detection of third parties in the camera's field of view, and a private chat with the teacher.

UC 3 — Conducting Lessons

The teacher and classmates' perspective. Includes starting and transferring lesson content, displaying transmitted data, and adjusting output volume for different contexts such as group work.

UC 4 — Robot Control & Spatial Privacy

The child navigates the robot remotely through the classroom — within the boundaries defined by no-go and no-view zones.

Conclusion

Privacy, designed well, changes what's possible.

PRIVATAR demonstrates that telepresence for children with long-term absence from school doesn't have to mean choosing between participation and protection. When privacy is treated as a core design principle, not a compliance layer. It becomes the very thing that makes genuine inclusion possible.

The project's findings point beyond the classroom. The UX patterns, methodological approaches, and role-based control frameworks developed here are transferable: to other vulnerable user groups, other privacy-sensitive contexts, and other health innovation challenges where trust is the prerequisite for adoption.

Protecting presence isn't a constraint on design. It's the brief.

Erschienen in

CERN IdeaSquare Journal of Experimental Innovation

Jahr

2026

Autoren

Philipp Schütz mit Oliver Gerstheimer & Isabelle Friedrich