A value-add process, not a new material.
Programmable magnetization doesn’t mine anything, sinter anything or replace NdFeB — it re-magnetizes the magnets the existing supply chain already makes. That single fact organizes the whole industry picture: the underlying markets are huge and measurable; the coded layer on top is early, niche, and dominated by one IP holder.
The markets coded magnets ride on
These figures describe the conventional magnet industry — the substrate. They are the only market numbers on this site, and each carries its named source.
Rare earths: the constraint that flatters coding
China performs roughly 69% of rare-earth mining and about 90% of downstream processing and magnet-making (IEA). In April 2025 China placed export controls on dysprosium, terbium and NdFeB magnet products — turning magnet supply from a procurement line-item into a board-level risk for every OEM that ships motors, sensors or speakers.
Two consequences favor engineered magnetization. First, force per gram: a pattern that concentrates flux where the work happens extracts more useful force from the same material — coding stretches scarce NdFeB rather than consuming more of it (it redistributes energy; it cannot exceed the material’s ~52 MGOe energy product). Second, the magnetizing step is not China-dominated: printing patterns is a machine-and-software business, not a mining business, and the machines are built by Western and Japanese magnetizer firms. The value-add layer sits outside the choke point.
Who else is in this territory
Correlated Magnetics Research
Inventor of correlation-coded magnetization; an IP fortress — cornerstone patent US 7,800,471, 121 issued + 58 pending by end-2014 (“100+ US” per its FAQ). Sole primary source of true coded parts (Polymagnet®), with Industrial Magnetics Inc. as verified licensee/partner (Max-Attach®) and retail via Amazing Magnets and PolarStar. Note: the earliest 2008-priority patents expire around 2028–2030 — a date that could open the field.
Magswitch & electropermanent makers
Switchable and electropermanent magnets are “programmable” on the time axis — on/off, no standby power — and they are shipping at industrial scale. Coded magnets are programmable on the space axis and add information (identity, alignment). The two solve adjacent problems and increasingly appear in the same cells.
Encoder rings, fridge strips, Halbach houses
Simple periodic multipole magnetization predates CMR entirely (prior art to US 3,127,544): refrigerator-strip magnets, ABS encoder rings. Halbach producers (Integrated Magnetics, EPI) build the flux-focusing end. Chinese houses (SDM, MPCO, HSMAG) sell lookalike multipole parts — whether any true correlation-coded copying occurs is unverified.
Apple — and the academic frontier
Apple holds its own coded-magnetic-connector patents (US 2012/0021619) — big OEMs can design around or license as they choose. Meanwhile the academic frontier is printing programmable magnetization into soft materials: shape-morphing elastomer microrobots (MIT, Nature 2018; Nature Communications 2024–25) — magnetization patterns as the actuators of machines with no rigid parts at all.
Adoption: what pushes, what blocks
- Miniaturization & wearables — force in less volume, field that doesn’t leak into sensors.
- Tool-less assembly & modularity — right-to-repair pressure rewards parts that align, latch and release by design.
- Field containment as safety — magnets acceptable near cards, compasses, implants and instruments.
- Force per gram — rare-earth economics reward extracting more work from less NdFeB.
- Functions with no other implementation — twist-release, keyed identity: no spring-and-catch equivalent exists in one passive part.
- Awareness spikes — a single SmarterEveryDay video put coded magnets in front of tens of millions of engineers-to-be.
- The physics tax — shallow reach; if your design needs force at distance, coding works against you.
- Cost & tooling — ~$45k desktop printers, NRE, $10–40 coded pairs vs <$2 plain magnets.
- Single-source IP — one primary supplier is a procurement red flag for high-volume programs.
- Design effort — a new CAD layer ($4,999/yr Polyvision) and a new intuition to learn.
- No field-strength miracle — same NdFeB energy; coding redistributes, marketing sometimes implies more.
- Thin catalog & sourcing — ~200 SKUs and no second source yet, versus a universe of commodity magnets.