What it is
A virtual laboratory built in Unity for the Quest 3, with an iPad console for the teacher and a web dashboard for results.
It covers molecular geometry, chirality and isomerism in chemistry; the Lorentz force, swinging conductors and induction in physics; and visualisations in complex mathematics. It's around 31,000 lines of code.
Why it exists
Stereochemistry is taught on paper using wedge-and-dash notation. A wedge means a bond points towards you, a dash means it points away, and from that you're expected to see that two molecules with identical connectivity are mirror images which cannot be laid on top of each other. Some people see it instantly. Most of my class didn't, including people who were better at chemistry than I was.
The difficult part is translating a three-dimensional structure from two-dimensional notation.
Students are decoding and learning at the same time. In VR you can hand someone the molecule, let them turn it, and put its mirror image next to it. The decoding step disappears and what's left is the actual concept.
How it works
The application runs natively on the Quest 3. Molecules are generated from a structural definition rather than modelled one by one, so adding a compound means adding data rather than geometry. Chirality is shown by generating the mirror image of whatever molecule the student is holding and letting them try, and fail, to superimpose the two.
The physics modules recalculate the outcome whenever a student changes a parameter. For example, the swinging-conductor demonstration computes the force on a current-carrying wire from the field strength, current and length of wire in the field.
The teacher's iPad console controls what the headset shows, which matters in a classroom where you can't see what a student is looking at. The web dashboard records task completion and timing for the study.
Study results
We ran a first round of testing with students at our school and compared their performance against conventional teaching of the same content. The results leaned in favour of the VR group, but the dataset was small and the effect wasn't large enough to conclude anything. We wrote that down as a tendency rather than a result, both in our submission and when we presented it.
The data did not support a clean positive result, even though that would have been easier to present at a competition.
The next round needs a larger sample, a cleaner control condition and interviews, because completion times on their own can't distinguish between students who understood more and students who simply enjoyed it more.
What I got wrong
We tried to cover too many subjects.
Chemistry, physics and mathematics modules across three subject areas, and 31,000 lines to maintain, while the study — which was the actual scientific question — ran on a small dataset collected late. If I did it again I'd build the chirality module alone, spend the remaining time on a proper experimental design, and have an answer instead of a tendency.
Where it stands
The application works and has been used in lessons. The study is unfinished. The project won its regional Jugend forscht competition in Munich South and went on to the Bavarian state round in April 2026, where 59 projects were presented by 87 participants. It didn't advance to the national final.