What I'm focused on.

A short, deliberate list of where my attention is this month. Updated when it stops being true — not on a fixed schedule.

Last updated · 4 May 2026 · London

This page follows the /now convention started by Derek Sivers. The idea: a single canonical answer to "what are you up to these days?" that's actually current, instead of a year-old bio. If you've landed here from an email or DM, this is the most up-to-date snapshot.

§ Building

AcoustR — turning microphones into condition-assessment instruments.

Pre-product. Right now I'm working through the signal-to-decision pipeline that lets a single microphone classify the health state of an engine, gearbox, or rotating asset — and packaging it so a non-specialist can read the output. The maths is built; the work this quarter is on calibration, edge-case data collection, and the deployment story.

Field-data collection trips with two pilot operators
Going from waveform → diagnostic class with an interpretable middle layer
A first hardware reference design that ships with the model

Public site: acoustr.com. The on-site venture page is at /ventures/acoustr.

§ Researching

Extending NDCI beyond the doctoral scope.

The published NDCI work covered Engine, Fuel, EPS, and ECS. I'm reading and writing toward two extensions: (1) NDCI under non-stationary degradation modes — i.e. when the diagnostic landscape shifts mid-mission — and (2) lifting it from sensor-selection to a general decision-quality metric for any inference graph.

Re-reading the diagnostic-Bayesian-network literature with a fresh eye
Drafting a follow-up paper outline with the Cranfield co-authors
The framework lives at /mosof and /ndci in plain English

§ Writing

Site-native essays from the doctoral work.

Rather than backfill old material, I'm writing the on-site explainers fresh. Each piece tries to take one method and make it readable for the person who will actually deploy it — fleet engineer, ops director, technical investor — not the person grading it.

Subscribe via the Atom feed if that's how you read
Listing of what's out there at /writing

§ Consulting

Selectively open for new work.

1–2 engagements this quarter

Where I add the most leverage is the first 2–4 weeks of a sensor / diagnostic / optimisation programme — the brief, the framing, the first cut at the trade-off space — before the implementation team scales up. Tier outline and specifics on the /consulting page.

Preferred: aerospace, energy, heavy industry, regulated medtech
Open to advisory / fractional CTO conversations with deep-tech founders
Reach out: [email protected]

§ Reading

What's on the desk.

Selected NSGA-III / MOEA-D papers — the state-of-the-art still has texture I want to internalise
Re-reading The Design of Everyday Things with a "cockpit UX" lens
Backlog: anything sharp on building deep-tech ventures out of doctoral work