The bottleneck in research isn't ideas.
It's how fast you can hold one in your hands.

Regulators set a clock.

• US CongressThe FDA Modernization Act 2.0, signed December 2022, ended the standing requirement that a drug be animal tested before human trials and names organoids and organs on chips among the alternatives. It permits rather than requires, so it opens a door rather than closing one.
• FDAMoving to replace animal testing with organ on chip systems and organoids within five years.
• NIHStopped issuing funding opportunities built solely around animal models.
• EU ParliamentVoted in September 2021 for an action plan to phase out animal use in research, testing and education, targeting full replacement as soon as it is scientifically possible.
• EU CommissionAdopted a roadmap to phase out animal testing for chemical safety on 1 June 2026, with more than thirty recommendations spanning chemicals, pharmaceuticals and food additives.
• UKCommitted to the same path, with £75 million behind it.
• EPASet a 2035 deadline to eliminate mammal testing.

Capacity did not move with demand.

Microfluidic chips are still cast by hand. Soft lithography needs a cleanroom to build the mold, then, in rough order of what they cost, a mask aligner, a wafer bonder, an automated coater track, a mask exposer, a plasma system, and a spin coater, plus a vacuum desiccator, a developer bench, hot plates, a curing oven, a profilometer to check what you actually made, and the fume hood and consumables around all of it. It needs someone trained to run the whole chain. That is why the capability lives inside a small number of institutions.

The process itself is slow and operator dependent, with cure conditions that shift batch to batch. And labs prototype in PDMS while manufacturing in thermoplastic, so the device they validate is never the device they ship.

The demand runs wider than biology, and it has to scale with AI.

Regulation set the clock for organ on chip work, but it is not the only pressure. When Microsystems and Nanoengineering surveyed scholars and industry for its ten ground challenges for the field, fabrication ran through the answers. Several of the ten stall not on ideas but on access to fast, high fidelity, biocompatible microfabrication: functional organ on chip reconstruction, affordable point of care diagnostics, precision that survives the print in high performance devices, and the AI guided design to fabrication loop that only closes when iteration is cheap.

Every one of those sits on the same microfluidic and additive foundation we build for. A process that turns a standard printer into a source of clear, biocompatible parts in minutes moves the fabrication half of each problem at once. The pull is not a single market. It is a field wide shift toward tools that let a lab hold an idea in its hands the week it has it, across diagnostics, computing, and medicine alike.

Underneath all of them sits the same dependency: fabrication fast enough to prototype, cheap enough to repeat, consistent enough to automate, and able to scale without switching process. Where our process fits is the rest of this page. How we intend to commercialize it is deliberately not public, and we share that under NDA with investors and partners who need it.

The market is large, and the analysts disagree about how large.

Eight research firms publish a microfluidics market size. For nominally the same market they report figures between $9.4 billion and $41.9 billion, a spread of more than four times. That gap is about scope, whether a report counts diagnostics alone or the full device, component, and instrumentation ecosystem, rather than about growth.

What the eight agree on matters more than what they dispute. All of them project sustained growth, North America leading, and medical and diagnostics as the largest application. Six of the eight put compound growth in double digits, and the two lowest are 8.3 and 8.7 percent. Mordor Intelligence carves out organ on chip as the fastest growing sub technology in the field at 22.7 percent, which is the segment our biocompatible housings serve.

Growth at those rates runs straight into the constraint described above. A market cannot compound at double digits for a decade while the parts it depends on are still cast by hand in a cleanroom. That is the bottleneck, and it is the one we removed.

What the alternatives cost.

Every existing option forces a tradeoff between speed and scale. Soft lithography iterates but cannot produce. Molding produces but cannot iterate. We hold both, in one material, with no cleanroom and no tooling.

Where Aionik sits in the chip fabrication ecosystem, route by route →

Most labs own none of this, so they send the design out and buy finished chips. That is not a fabrication method, it is a way of paying for one, which is why it is not in the table. It removes the capital cost and replaces it with a queue: one to twelve weeks per revision, priced per order, with setup and tooling charged against every new design.

The tradeoff is also a bill. A research bench that cannot produce at volume starts at $65,000. A line that can runs $650,000 to $1.67 million in equipment, and the cleanroom the mold step needs costs $250 to $1,000 per square foot to build, depending on how fine the features are, then carries an operating cost every year it stays in service, in energy, filters, maintenance, and recertification.

See how each number is built, and the full process for both PDMS and injection molding, with sources →

What a cleanroom is, which industries need one, and what it costs →