The University of the Arts London this week opened a facility combining wearable brain imaging and AI analytics to ask why a slow, dialogue driven cartoon like Bluey and a fast cut, action spliced show like PAW Patrol may not feel the same to a preschooler.
Why does a slow, dialogue-heavy cartoon like Bluey feel different to a three-year-old than the fast-cut, action-spliced world of PAW Patrol? A new research lab in London is betting that the answer lies not in counting screen-time minutes, but in measuring what specific shows actually do to young brains.
The University of the Arts London opened the UK's first dedicated media-neuroscience facility, called the Nerve Lab, the week of 13 June 2026, combining wearable brain imaging, motion capture, and AI-powered analytics to study how people respond to media and artistic experiences in real time. One of its first projects, Animating Minds, is focused squarely on children's content.
Prof Tim Smith, who directs the lab, frames the central question simply: short-form, fast-paced, spliced programming may shape attention, comprehension, and emotional response in ways that are quite different from slower, dialogue-driven shows. "It's not just about minutes," he told The Guardian. "It's about what the content does."
That framing is what animates the Animating Minds team. Research assistant Alisa Musatova says she keeps meeting parents whose children, as young as two, are spending three or four hours a day on screens, while researchers still do not know what is appropriate for which age. Her group has built a database of roughly 1,000 episodes of popular animated television, then deployed AI tools to analyse pacing, colourfulness, loudness, shot frequency, and narrative structure. The numbers are paired with interviews of animators, producers, and commissioners, and the team is currently recruiting UK families with children aged three to six for an online study on short-term attention.
A second project, Mathstronauts, takes the question into a north London primary school. Dr Rakhi Leela Nair uses functional near-infrared spectroscopy, a wearable neoprene sensor cap that reads brain activity by measuring blood-flow changes, while seven- and eight-year-olds play a fractions game. The system adapts in real time: if a child keeps making errors because they are jumping in too fast, it shifts to a slow-down, inhibition-training mode. If the errors look like genuine conceptual gaps, it shifts to teaching the concept. "We're trying to see what the brain does when maths gets hard," Nair said.
Independent researchers welcome the method but flag what it can and cannot show. Prof Heather Kirkorian, a developmental psychologist at the University of Wisconsin-Madison, says precise measurement of children's media effects has long been sparse, and AI-based content analysis finally enables the kind of scale that manual coding cannot. Polly Conway of Common Sense Media calls quantification of features like developmental pacing "really useful." But Prof Roi Cohen Kadosh, a cognitive neuroscientist at the University of Surrey, cautions that the value of the maths brain-imaging work depends on whether it adds insight beyond what teachers and standard assessments already reveal.
The methodological limits are real. Wearable brain imaging captures correlates of attention, not long-term developmental effects. A database of 1,000 episodes is a starting corpus, not a population. And readings from a sensor cap on a wiggling seven-year-old are noisy. The lab is positioned as early-stage research infrastructure, not a verdict engine. As Smith puts it, the goal is to give parents, teachers, and content makers a sharper vocabulary for asking what a show is actually doing, not to declare which ones are safe.
What does that vocabulary look like in practice? For a parent choosing between a Bluey episode and a PAW Patrol marathon, the questions the lab's framing suggests are concrete: how many cuts per minute, how much dialogue versus action, whether the storytelling holds a single idea long enough for a child to follow it, and whether the pace leaves room for emotional read. For a teacher, the Mathstronauts work hints at adaptive tools that could distinguish impulsivity from confusion in real time, though the research is years from any classroom product. For a content maker, a database of 1,000 episodes, tagged and quantified, is a feedback loop that did not exist a decade ago.
The lab is also reaching beyond children's programming. Other projects include tools to help visually impaired people navigate video games and systems that shape live dance and music performances using brain and behaviour data. The throughline is the same. Instead of treating media as a single variable to be minimised or maximised, treat specific features as testable.
That is the practical stake. The old question, "how much screen time is okay?", treats every minute as fungible. The new question, "what is this show doing to attention, comprehension, and emotion in this child at this age?", does not. Until the research matures, the honest answer for most parents is closer to "watch the show with your child and notice the cuts" than to any number of minutes. The lab's contribution, if it delivers, will be to replace that advice with evidence.