Research, shipped.
Methods only matter when they meet a real machine, a real signal, and a real customer. The ventures below take ideas from the thesis and the lab — sensor optimisation, diagnostic-contribution scoring, multi-objective trade-offs — and turn them into products and tools.
Founder thesis
The most expensive engineering decisions are not "what algorithm" — they're "what to measure, what to trust, and what to do next". My ventures address each of those three, starting with acoustic diagnostics for engines and machines.
Three ventures. One thesis.
AcoustR
A product application of the sensor-optimisation and diagnostic-contribution methods from the thesis. AcoustR turns microphone signals into condition assessments — letting operators detect faults, monitor wear, and triage maintenance from sound alone, without the cost or installation overhead of a full vibration-sensor stack.
Sensorry
An applied tool that packages the MOSOF / NDCI workflow for engineering teams designing or auditing sensor networks. Where AcoustR ships a diagnostic, Sensorry ships the design step that comes before it: which sensors, where, and at what cost-coverage trade-off. Currently in prototype.
CompAeros
An early concept extending diagnostic-contribution thinking to composite-structure health on aerospace airframes. Not yet public — listed here for transparency about the long-range portfolio rather than as a current product.
Building something diagnostic?
If you're an investor exploring the diagnostic-AI / industrial-AI space, or an operator with a hard signal-to-decision problem you'd consider piloting, AcoustR is the lead venture and the right entry point.
Request the investor memo →