INVENT
Design new low-voltage diagnostic tools or software when an existing capability isn't enough.
CONCEPT / REQUIRES TESTING ↗Seven real projects growing out of hands-on engineering experience and self-directed, AI-assisted development. From fabrication CAD and press-brake planning to local AI and experimental game worlds.
Not a list of invented launches: explore the actual previews, measured advances, remaining tests and bigger ideas behind the builds.
Prototype → test → improve → publish
ACTUAL DEVELOPMENT PREVIEW · 20 OCT 2026From an operator-led interface and editable profile library to 2D/3D models, reference developments, DXF work and printable A3 engineering sheets. This has become a substantial CAD engineering programme.
Manufacturing-safe unfolding and workshop validation remain separate approval gates — not a percentage score.
Explore the Norman project ↗Projects developed from real manufacturing know-how, independent learning and persistent AI-assisted engineering — with pictures of the work.
OPEN A PROJECT FOR IMAGES, MILESTONES AND ROADMAP ↗This is the public development diary. New results appear here when an update is reviewed and published — with successes, unfinished work and blockers distinguished.
APPROVED UPDATES / NOT PRIVATE LIVE LOGSDevelopment milestones, tests and decisions — shared openly as progress is reviewed.
Five long-range Nigel AI ideas exploring how computers could extend their capabilities through physical tools, simulations and connected devices.
RESEARCH CONCEPTS / NOT SHIPPED FEATURESDesign new low-voltage diagnostic tools or software when an existing capability isn't enough.
CONCEPT / REQUIRES TESTING ↗Reconstruct failure timelines from logs, sensor readings and other evidence.
CONCEPT / REQUIRES TESTING ↗Rehearse changes in isolated simulations before risking working equipment.
CONCEPT / REQUIRES TESTING ↗Bring authorised devices, instruments and computers into one investigation.
CONCEPT / REQUIRES TESTING ↗Build understanding of unfamiliar equipment from lawful evidence and measurements.
CONCEPT / REQUIRES TESTING ↗One Nigel AI, expandable abilities. The long-term goal is a capability engine that can identify the software, interfaces and instruments needed for a task — not pretend to measure or control equipment it isn't connected to.
From a working Windows assistant to self-repair, portable diagnostics and ambitious connected engineering. These are planning windows, not promised delivery dates.
WORKING → TESTED → PLANNED → RESEARCHTimeline is a forward-looking estimate for sustained development, not a commitment. No autonomous hazardous repair or machine control is proposed without qualified human review.
Manufacturing experience informs the requirements for geometry, bend sequencing and operator decisions.
Prototypes, test counts, working demonstrations and real-world acceptance are clearly distinguished.
Local processing, human approvals, resilient updates and portable software are priorities as the products mature.
Years spent around press brakes, turret punching, fabrication and technical problem-solving shape these independent software projects. Without a conventional software-development background, this independent work has grown through hands-on learning and AI-assisted programming, backed by practical engineering judgement, repeated tests and a determination to make tools people can actually use.
Seven products spanning AI, industrial CAD, bending, financial research, web ordering and game simulation demonstrate more than one technical skill: they show initiative, persistence and an ability to turn everyday problems into working prototypes. This is a public record for collaborators, future customers and employers.
Newly approved milestones will appear in the build journal. Commercial releases, demos and contact information will be announced when ready.