Biological systems are noisy, individual and rarely cooperative. The gap between a measurement that is technically correct and one that is clinically useful is where most of the engineering effort goes, and it is the part of the problem we treat as central rather than incidental.
What we work on
Our work in this area spans the full path from a physical phenomenon to a decision: the transducer that senses it, the analogue front end that conditions it, the acquisition chain that digitises it without destroying what matters, and the analysis that turns the result into something a clinician or researcher can act on.
That full-path view matters because the failure modes cross boundaries. A filter chosen for a clean bench signal can quietly remove the feature a study depends on. A sampling decision made in firmware can put a ceiling on every analysis downstream of it. Splitting those decisions across separate teams tends to mean nobody owns the consequence.
How this work is constrained
Anything intended for clinical use carries obligations that shape the engineering from the first day rather than the last: design history, risk management, verification evidence, and a traceable line from requirement to test. We build to those obligations where a programme calls for them, and we say plainly when a piece of work is research rather than a regulated deliverable. Conflating the two helps nobody.
Where work touches human subjects or identifiable data, ethical approval and data protection are preconditions of the work, not paperwork appended to it.
Working with us
Programmes in this area usually begin with a measurement problem rather than a product specification — a phenomenon someone needs to observe reliably, at a cost and in a setting that existing instruments do not serve. If that describes your situation, get in touch.