Four programs. One public test.

A useful quantum computer cannot be a laboratory demonstration that holds coherence for a few microseconds. It has to survive thousands of hours of operation, tolerate component failures, and produce results that another group can reproduce. That is an architecture problem as much as a physics problem.

Our group works across the stack: superconducting transmon modules, cryogenic control and readout, quantum low-density parity-check codes, and decoders fast enough to keep up with the machine. We are particularly interested in modular designs, where many small processors are linked into a larger logical machine. Modules are easier to build, test, and replace — and they make failure modes visible.

The test of success is not a single record fidelity. It is a logical qubit whose error rate falls as the code distance grows, measured repeatedly, on hardware that a graduate student can calibrate without a specialist.

Quantum sensing is often treated as the practical little sibling of quantum computing. We think it deserves the front seat. A magnetometer that can map groundwater recharge, a compact clock that keeps a grid synchronized, or a sensor that detects a contaminant at parts-per-trillion can matter to a public agency this decade, not next.

Our sensing group builds field instruments with explicit tradeoffs: sensitivity against portability, automation against repairability, and precision against the cost of calibration. We test in real conditions — on farms, in utility substations, and in partner clinics — because a sensor that only works in a basement is not yet a public capability.

Every design file, calibration routine, and failure log is released under an open hardware license. The point is not to sell instruments. The point is to lower the cost of knowing.