Designing for AR: Lessons from a 1-day prototype

3 simple lessons learned testing a prototype educational Augmented Reality experience for BBC Bitesize.


Floating landscape with mountain, river and clouds
It’s been almost 20 years since BBC Bitesize launched to help students across the UK learn and revise in small, bite-sized chunks.

The BBC Education team recently undertook a Design Sprint to explore how Bitesize might utilise new technologies into its third decade. One of the outputs of that frantic week was a mid-fi Augmented Reality (AR) prototype to introduce Key Stage 3 Geography students to the Water Cycle.

Our starting point was this study guide. More specifically, we looked at the accompanying graphic for inspiration:

On the fourth day of the sprint, we used the VR app Google Tilt Brush to knock up a quick model , before importing it into Unity to add labels and content (as well as some pretty animations 😁). We almost entirely ignored onboarding – we wanted to see what people knew already without any guidance.

On the final day of the sprint, we took the prototype to editorial staff from BBC Learning and BBC Children’s to gain some general feedback. Although we would have loved to test it with appropriately aged students, it just wasn’t feasible at the time. With this in mind, all of this learning could be completely diferent when it comes to younger users.

What did we find out?

People don't know how to "start" AR yet.

Only 1 (of 10) participants understood that their phone would have to detect the floor before they could place the model in the world. This involves moving the phone slowly around the environment so the camera can pick up “feature” points on the floor, before joining them up to make a horizontal plane.

Users need to detect the floor before they can place a digital object “on” it

People stand still, unless you tell them to move

Only 2 of our participants started to walk around their physical environment to explore the digital model they had just dropped into it. The vast majority rotated on the spot and tried standard phone interactions (tap, pinch/zoom) to try and manipulate the landscape. We deliberately excluded screen interactions for this prototype as we wanted to explore the hypothesis that moving around the model would enhance learning outcomes.

"Hotspots" became visible as users walked towards them

Exploring "underneath" isn't natural

Once we had users moving to explore, the final part of the experience has them go underneath the mountain to follow the water percolating back to the sea. Although there was a faint path for users to follow, none of them considered crouching to go inside and under the mountain.

Users didn't realise they could explore the world underground

We were aware of the potential for some of these challenges going into the testing, thanks to some AR UX tips published by Google , but it was useful to gain some personal validation ourselves, especially within an educational context. It also tallies with lessons learned by our colleagues in BBC R&D from the BBC Civilisations AR app.

As with a lot of research, we came away with more questions than answers. We’d really like to explore whether an educational AR experience is best suited to the classroom or learning at home, as well as what areas of the curriculum are best complemented by Augmented Reality activities. Most importantly, we need to consider how we’d build the experience to include users of assistive technology and those with mobility issues.

At the same time, we need to temper ourselves and remember that the future of Augmented Reality is likely not through a phone or tablet held at arm’s length. Everything we learn in this "bridging" phase of AR technology could change in a flash if and when it becomes embedded into our everyday lives.

Special credit to Jonathan "Jops" Halkett for creating the asset and using his Unity skills to pull the team's vision together.

Where next?