Read the system from its boundaries
The work starts where responsibility becomes unclear: between teams, systems, contracts, policy positions, data flows, and real-world operational pressure.
Independent analytical systems consultant
Finding the gap between systems, ownership, evidence, and action.
I help organisations make sense of complex operational problems, define the control chain that is missing, and turn the finding into a decision-ready pack that can be implemented.
Pattern
Boundary failure
Two working systems meet, but nobody owns the handoff.
Output
Control design
Named ownership, timing rules, escalation paths, and evidence standards.
Use
Decision pack
Clear enough for boards, funders, regulators, buyers, and delivery teams.
Method
Analytical work built for handover, not theatre.
Separate evidence from assumption
Published facts, records, standards, interview evidence, failure history, and operational constraints are turned into a clean, defensible picture.
Design the missing control chain
The answer is expressed as ownership, minimum information, escalation rules, timings, oversight rhythm, and practical implementation steps.
Work areas
Cross-sector analysis where the common problem is coordination failure.
Health systems
Discharge flow and bed release control
Scope notes, stakeholder interview guides, gap analysis frameworks, target operating models, and 90-day implementation handover packs for hospital flow problems.
Water and utilities
Incident control and supply resilience
Buyer cases and operating reviews for resilience failures, root-cause ownership, minimum incident records, escalation clocks, and exception-review routines.
Defence and security
Capability concepts and procurement packs
Open-source analytical submissions, technical briefs, cost models, programme execution plans, threat assessments, and structured funding proposals.
Policy and markets
Regulatory logic and legislative analysis
Executive briefs, crosswalks, hostile-read analysis, and plain-English policy redesign focused on where classification does not match customer harm.
Selected analysis
Four pieces of real work, described by problem and method.
Organisations, programmes, and commercial detail are not published. What is shown is the reasoning — which is the part that transfers.
Autonomy and robotics · sensing
The sensor limit nobody specifies
- The problem
- Sensors are specified by what they can see. Control loops are limited by when the sensor can tell you. The second constraint is rarely written down anywhere.
- The analysis
- Took the detector time constant a manufacturer publishes, derived the channel's corner frequency from it, and set that against the Nyquist limit its own frame rate already permits — showing usable bandwidth is being discarded after it has been sampled. Then read the existing academic work closely enough to establish exactly where it stops short.
- The output
- A bounded claim with the unrecoverable part stated explicitly, an experiment scored against physical ground truth rather than appearance, and a written list of the three results that would disprove it.
Air defence · interceptor sensing
A requirement re-specified by its latency budget
- The problem
- A published requirement to hold a manoeuvre within three metres of a target at near-Mach closing speed, framed by everyone as a detection-range and discrimination problem.
- The analysis
- Divided the two numbers the requirement itself states: at that closing speed, three metres is under nine milliseconds — less than one frame period of the sensor normally proposed, before any processing has run. The specification was rebuilt around an end-to-end latency budget, with each channel assigned the job its physics can actually do.
- The output
- Full technical pack: specification, latency budget, comparative analysis, costed programme plan, and a section setting out where the approach may fail.
Maritime autonomy · humanitarian survey
Costing the operation, not the sensor
- The problem
- Surveying contaminated waterways for submerged ordnance. A cheap sensor is not a cheap solution if it needs a boat, a crew, and a day rate — and in many of the affected places there is no boat to hire at any price.
- The analysis
- Costed the whole operating chain rather than the instrument, sized detection performance against published magnetometry literature and manufacturer datasheets, and held the design to mature open-source autopilot firmware so nothing rested on unproven software.
- The output
- Complete system specification at a fraction of the stated cost ceiling, with standards-compatible survey output, no personnel on contaminated water, and every limitation declared in the document rather than discovered later.
Deployed medical · logistics and traceability
The gap was the release decision, not the test
- The problem
- Emergency transfusion in austere conditions, where decisions default to one blood type because the supporting information is not available at the point of need.
- The analysis
- Established that the diagnostic components were already validated and available — so the missing element was not an assay but the combination of typing, screening, release logic, and traceability in a single format a non-specialist can operate under pressure.
- The output
- A release protocol with four defined outcomes, a paper traceability chain readable without system access, operator error controls, and a costed validation programme.
Submissions
Full documents, where publication is useful and appropriate.
These are included to show the depth of the work: structured problem framing, evidence discipline, declared limitations, programme design, and handover-ready writing.
United Nations Development Programme challenge · March 2026
Community Underwater Ordnance Mapper
Autonomous non-technical survey concept prepared for a United Nations Development Programme challenge on underwater explosive ordnance. The design is built around community use, low capital cost, shore-based operation, mature open-source autopilot firmware, and IMSMA-compatible suspect-area mapping.
- Challenge
- Affordable tools to find and mark underwater explosive ordnance
- Status
- Advanced to final detailed evaluation stage; not selected for award
- Public record
- Public summary of the submitted concept, including the problem, system logic, declared limits, validation pathway, and outcome
Representative outputs
Documents that move from diagnosis to action.
The work usually lands as a complete pack: scope note, buyer case, operating model, evidence framework, stakeholder guide, implementation handover, and executive brief.
Buyer case
Gap analysis framework
Target operating model
Stakeholder interview guide
Implementation handover pack
Executive brief
Threat assessment
Procurement readiness plan
Engagements
Focused support when the problem is real but the ownership is unclear.
Rapid analytical review
A short, evidence-led review that defines the actual failure mode, the accountable boundary, and the decision that needs to be made.
Control-chain design
Practical operating design for ownership, escalation, timing, minimum information, exception review, and management oversight.
Bid, buyer, and board packs
Structured cases for funders, public buyers, regulators, executive teams, and senior decision-makers who need clarity without operational noise.
About
I find the joins.
My work is looking at two systems that both work, finding the boundary between them that nobody owns, and characterising it well enough that somebody can act on it.
I have done that across print procurement, mechanical engineering, and supply chain, and before that as a firearms officer with West Midlands Police — where the job was reading an incomplete situation quickly and picking the one variable that mattered. That habit does not leave you.
The discipline that came with it is the part clients actually buy. Every external claim is cited to a published source and checked before it leaves the desk. Assumptions are labelled as assumptions. Where the evidence does not support the conclusion someone was hoping for, the document says so — and says what would have to be true for it to change.
- Practice
- Independent. Sole practitioner. No vendor relationships or referral fees.
- Based
- United Kingdom. Work delivered remotely, in writing.
- Material
- Open sources and material you provide. No classified work undertaken.
- Confidentiality
- Client identities and engagement contents are never published or used as marketing.
Contact
Bring the problem, the documents, and the pressure point.
If the system is producing delay, risk, wasted effort, or regulatory exposure, the first step is a concise conversation about the boundary that is failing.