Explorers Build
Explorers Build
Five Explorers. Eight weeks. 25 iterations. And a working proof that hands-on engineering beats classroom physics.
This project began before Explorers Build existed. It's why Explorers Build exists.
In February 2026, a Y10 teacher at Kings College Cascais asked whether anyone in the Explorer's parent community would be willing to Captain a real engineering project with a small crew of students working toward a CREST Gold Award. One parent said yes.
The Explorers picked a topic that would have been laughed out of most STEM programmes: Beyblades. The pitch was simple. Beyblades are toys. But behind the toy is real rotational-dynamics physics — moment of inertia, angular momentum, friction models, the kind of engineering trade-offs that any spinning machine has to solve. If they could genuinely engineer a Beyblade that would spin dramatically longer than an off-the-shelf one, they'd have used exactly the physics and methodology of every serious rotating machine in the real world. Beyblades weren't the point. Building was the point.
Five students committed. One dropped off partway through. This is honest, and worth naming — real Crews aren't perfect, and losing an Explorer to competing priorities is part of the work. Explorers Build's design accounts for it now (small enough Crews that a departure doesn't collapse the project, structured hand-offs, and Captain 1:1s to spot risk early). Back in February, we learned this by living it.
Not simplified for kids. The actual theory.
The Captain ran two 90-minute physics sessions on rotational dynamics — the same content a university physics student would see in first-year mechanics, taught with kitchen-scale analogies and worked out on whiteboards until every Explorer could re-derive it.
In translation, a force accelerates a mass: F = ma. In rotation, a torque accelerates a rotational inertia:
where τ is torque (rotational force), α is angular acceleration, and I is rotational inertia — how much a spinning object resists changing its rotation. And here's the trick: I depends on where the mass sits. A ring of mass at the outer edge resists rotation much more than the same mass packed near the centre. This is why the Explorers ended up with a design where the outer ring carried most of the weight.
Angular momentum: L = I × ω. Rotational kinetic energy: K = ½ I ω². Once launched, a Beyblade can only lose energy through friction (at the tip touching the stadium) and air drag. Model those, and you can predict spin time before you build anything:
Higher rotational inertia → longer spin. Lower friction → longer spin. This became the design rule for every iteration.
Modular design meant they could test combinations, not just prototypes.
Instead of designing one Beyblade at a time, the Crew designed a modular Beyblade in Fusion 360 — three snap-together pieces: a top layer, a disk (the main mass), and a base (the tip that touches the stadium). Any disk could pair with any base. Any base with any disk. That meant five disk designs × five base designs = 25 combinations to test — and by holding one variable steady at a time, they could see exactly which change was doing the work.
They 3D-printed the parts in plastic (first at home, then at Lisbon FabLab for higher precision), logged every launch in a shared Google Sheet, and used simple statistics — medians, not averages — to keep results honest against the occasional imperfect launch.
One breakout combination. Two clean rules for spin.
Two patterns leap out of the map:
Rule 1 — Mass on the outside. Reading across each base row, the disks with more perimeter mass (Double Ring + Mass, Extend) beat the Minimal disk by 30–100%. This is moment of inertia at work: same mass, but concentrated at the rim, resists slowing dramatically better.
Rule 2 — A true, tiny contact point. Reading down the Flat-point column shows the biggest jump of any base. A base with a single, narrow contact point minimises friction with the stadium surface — and does so without sacrificing the balance that a round or long tip lose.
The winning combination — Double Ring + Mass disk × Flat-point base at 40.4g — put both rules together and delivered a 164-second median spin. Notes described it as "very stable" with "good life after death" (the technical term for a Beyblade that keeps rotating slowly even after starting to precess).
From the first Minimal-Flat baseline (51s) to the winning Double + Mass × Flat-point (164s), the Crew quadrupled spin time.
Every combination in the design space was tested and logged. Zero configurations were left un-explored.
All prototypes went from Fusion 360 CAD → Lisbon FabLab 3D printers → same-day test in the stadium. Iteration cycle: hours, not weeks.
Three progress updates, sent to the teacher over eight weeks.
Not a certificate. Real skills, applied to a real thing they made.
Fusion 360 well enough to build a modular assembly with swappable parts — the same tool used by professional mechanical engineers.
How to prep files for print, choose settings, recover from failed prints, work with a fablab, and understand which quality issues are printer-driven vs. design-driven.
Identifying independent, dependent, and controlled variables. Choosing medians over averages when outliers are inevitable. Recording qualitative notes alongside numbers.
Building a Google Sheet from scratch with pivot tables, heatmaps, and comparative charts — turning a pile of numbers into a story.
Rotational inertia, angular momentum, friction models, and spin-time prediction — theory linked directly to design decisions.
A CREST-Gold-standard report that turned eight weeks of build-notes into a paper an assessor could read cold and understand.
And two things that don't fit on a skills list: the experience of iterating past failure, and the taste of shipping something real. When Kavi hit sub-40 seconds on iteration 1, he could have quit. He didn't. Twenty-four iterations later, the same Explorer had built something four times better than his starting point — and could tell you exactly which two rules made the difference.
This project worked because of the model, not despite it.
The Beyblade project ran ad hoc — through parent-teacher goodwill, informal Sunday huddles at the Captain's house, a shared Google Sheet, and a Captain willing to write a physics curriculum from scratch on the fly. It worked. But it required an extraordinary amount of Captain-hours, coordination, and improvisation that most families and most schools cannot reproduce.
Explorers Build is what happens when you systematise this. Same Crew-of-3-to-5 model. Same Captain-led weekly rhythm. Same "we teach the actual physics" bar. But with a First Mate (our AI teaching assistant) picking up the between-session load. With a library of pre-built Expedition materials so Captains don't rewrite from scratch. With a documented curriculum that maps to IB outcomes so schools can accept the work. And with a Buy One, Help One programme so kids without access to a passionate parent-Captain still get in.
The Beyblade Crew was proof of concept. Everything since is the version that scales.
Join the waitlist. Tell us what they're curious about. We'll be in touch as we open a cohort near you.
Join the waitlist