C2CSI (Co-imagining robots for enhancing Child-to-Child Social Interactions)

The project was funded through the University of Bristol's School of Engineering, Mathematics and Technology's Pump Priming scheme, which provides up to £5,000 to help early-career researchers initiate new collaborative projects. The multidisciplinary team comprised investigators from across disciplines, supported by a mentor and amazing collaborators, bringing together diverse expertise to tackle this design challenge.

The Workshop Experience

The research team conducted an interactive workshop at the Bristol Digital Futures Institute's Neutral Lab with 20 children and their guardians, alongside 5 researchers. The workshop was structured in four phases: exploring current technologies, reflecting on missing functionalities, designing new robots, and evaluating potential risks.

Phase 1: Familiarising with Existing Robots

We asked the children to interact with five different robot types to learn what kinds of robots exist already:

Mobile robots (Gilobaby): Voice-controlled robots that could move around and play music, responding to commands like "play some music," "go forward," or "dance."
Soft robots (Purrble & Qoobo): Therapeutic cuddly robots with responsive heartbeats, tactile feedback, and emotional expression through purring and tail movements.
Desk robots (Eilik & EMO): Small humanoid companions that showed emotions, danced, responded to touch, and could move freely on surfaces.

What Children Learned

We asked the children to reflect on playing with each kind of robot. They recognised that robots could serve as icebreakers and that the soft robots with limited interactivity might not help with making friends. Some noted that robots with human-like expressions and responses to touch and speech were most engaging. Interestingly, children also observed that robots could help shy children by introducing them to other children.

We identified key preferences: children favoured robots with 'human-like' capabilities, emotional expression, mobility for interaction with peers, and responsiveness to touch and gestures rather than touch screens. Conversely, stationary soft robots without faces were seen as less valuable for social connection and a little uncanny.

Phase 2: Crafting Innovative Robots

In the second half of the workshop, children designed their own robots using building blocks, craft materials, textiles, and paper.

Children's Designs

One child designed a robot that could cook for hungry friends while they played, with a head featuring multiple ovens and a face screen. Another created a mobile robot that could move around and communicate with people. A third designed a calm, portable companion for shy children that they could stroke and sit with that could also speak to other people on their behalf and build confidence. Another creation featured emotional expression through screens and gestures, alongside mobility and a portable design.

Safety Considerations

Children also thoughtfully evaluated potential safety concerns. They recognised that soft robots were safe because they resembled dogs without the mess, while mobile robots could potentially be tripping hazards or sneak around unexpectedly. One child said their flying robot might occasionally jump off a cliff "because it likes to fly" but reassured that it would never be dangerous for people. A few of the children designed robots with built-in cannons, but reassured us they were just for looking cool and wouldn't actually hurt anyone.

Key Findings from the Design Phase

Children were significantly more engaged with crafting materials than preset options (e.g. Lego, Knex). Their designs emphasised robots that could move and interact with other children, suggesting that movement and interactivity are crucial for social facilitation. Many robots incorporated 'helper' functions so friends could focus on playing together. Emotional expression through gestures and varied movements was a consistent feature. The main focus of most of the children was that, overall, their robot was really cool.

Project Impact and Next Steps

The research generated valuable insights for designing technology that genuinely enhances rather than replaces child-to-child interactions. Children valued robots that would interest other children to come and play with them.

The project has also extended to be part of a six-month installation at the WTC, with plans for qualitative analysis of the workshop recordings and responses: https://www.wethecurious.org/whats-on/events/research-island-rescue. The team is open to new collaborations and is exploring future funding opportunities to continue this important work.

Key Takeaway

This project demonstrates that when children are given agency in the design process, they create robots that prioritise interaction, movement, emotional expression, and practical helpfulness, all features that support rather than hinder genuine social connection among peers.

Illustration of the session by the amazing Camille: https://www.camilleaubry.com/scribing