01 / HOLD → TWIST → RELEASE
The magnet that lets go
FIXED
HOLD
It clamps down with real force — then you twist it a few degrees and it simply lets go. The pole code aligns one way and cancels the next.
hover to twist
Ordinary magnets only pull or push. Programmable magnets are printed with hundreds of tiny coded poles on a single face — so they self-align, spring, latch, and release with a twist. Not new materials. A new language for magnetism.
Each of these is one magnet, doing something magnets simply don't. Hover or tap a card to watch the rule break.
It clamps down with real force — then you twist it a few degrees and it simply lets go. The pole code aligns one way and cancels the next.
Pull it in and it attracts — until it crosses a set gap, where it pushes back and floats. A spring with no metal coil, tuned entirely in the pattern.
Bring two coded halves near and they rotate themselves into exact alignment — every time, one orientation only. Blind-mate connectors that can't go in wrong.
Detents you can feel — the satisfying click of a good dial — built from magnetic poles instead of springs and notches. No wear, no parts to break.
Larry Fullerton spent his career teaching radios to whisper — a pioneer of ultra-wideband, where information hides in precisely coded pulses instead of a single steady tone.
Then a simple wish nagged at him: he wanted to build a self-assembling toy for his grandchildren — pieces that would find their own place and snap together on their own. Ordinary magnets couldn't do it. They only know one trick.
So he asked a question no one had really asked: what if you could code a magnet the way you code a signal? Print not one pole, but a pattern of hundreds of tiny poles — maxels — arranged so they only lock with their matching mate, and cancel with everything else.
In 2008, in Huntsville, Alabama, he founded Correlated Magnetics Research to build it. What came out is called the first fundamental invention in magnetics since electromagnetism was discovered in the 1830s — the programmable magnet.
Larry Fullerton · 1950–2016 · the toy became a new kind of magnet
The last time magnetism gained a genuinely new chapter — until now.
Fullerton's ultra-wideband work: information carried in coded pulses, not raw power.
Pieces that find their place — a grandfather's project that ordinary magnets couldn't solve.
Signal-coding theory, applied to magnetism. The maxel is born.
Self-align, spring, latch, release — designed in software, printed onto one face.
The same coded-pole idea scales from a phone on your dashboard to a sensor on a factory line — quieter fields, cleaner snaps, releases on purpose.
Reach toward the dash and it pulls the phone into the same exact spot — then twists free when you want it back.
A screen that mates to its base one way only, holds firm as a tablet, and lifts away without a fight.
Bands and chargers that align themselves and release clean.
Modules that key into place with almost no stray field to interfere.
Fixtures, couplers and tooling that hold hard and let go on command.
Every coded behavior starts as an application note — the working papers where magnetism gets designed like software. Read how the patterns are built, measured, and put to work.
Join the field notes and we'll send the strangest, most useful things coded magnets can do — new behaviors, design ideas, and where to feel it in your own hands. No noise. Real dispatches.
The field notes are on their way. Watch for the first dispatch — and if you can't wait, go feel it in the demo kit.
Feel it yourself · demo kit →