Keir Archibald

Developed what is believed to be the first concealed electromagnetic drive system integrated into a UCI-class racing bicycle, built to deliver real assist while remaining undetectable under most current UCI inspection methods. No institutional support, no GCN technical resource, no workshop, £500 budget. Coordinated deliverables and produced build documentation for GCN’s media team while remaining sole technical lead from first concept to rideable prototype. Met the 20 W design target. Featured in a GCN video exceeding 500,000 views within 48 hours, prompting a second collaboration with Be Amazed (13M subscribers) and renewed engagement with the UCI on motor doping detection.

• Optimised magnet configuration using FEA in FEMM linked to a genetic algorithm, optimising torque-to-mass ratio, air gap, and runtime penalty, converging on 48 N52 magnets.
• Reverse engineered the fork via photogrammetry - 300+ images through Polycam to Fusion 360 mesh - and the wheel through physical measurement; delivered the integrated shroud to a production fit on carbon components in two iterations.
• Tested four coil wire gauges against resistance, mass, and thermal behaviour; redesigned around an aluminium heatsink core, extending time-to-thermal-limit by ~40%; a calibrated thermistor LUT then allowed current and duty cycle to be pushed to material limits, gaining ~10% additional output on the 20 W target.
• Wrote the full embedded control system from scratch on an ESP32-S3; position-synchronised firing and closed-loop PI current control at 1 kHz were critical to viability, without both, the coils either underperform or fail.
• Before the build, I travelled to Budapest to interview Stefano Varjas, the only publicly known motor doping mechanic, alongside Chris Marshall-Bell and Nicholas Raudenski (UCI anti-motor-doping lead), using those conversations to directly inform the system design.

Placement across a full-capability medical design consultancy: prototyping, machining, human factors, and industrial design. Direct responsibility on live client projects from day one within a small, senior engineering team.

• Designed the full mechanical methodology for a human factors dose counter tolerance study in collaboration with Rebus Engineering. Machined custom positional offset components and validated accuracy using a Cognex In-Sight vision system with adapted C++ inspection code, producing data used directly as regulatory evidence for medical trial approval.
• Built and used custom modular test rigs to characterise a biosimilar autoinjector device in a clean bio lab. Converting the test data into a full system requirements specification that formed the foundation of an in-house device development programme.
• Designed a novel latch mechanism from requirements through to a working 3D printed prototype, engineered to avoid patent conflicts with existing devices - mechanism entered early-stage patenting at the close of the placement.
• Built in-house manufacturing rigs: ultrasonic welding fixtures, a syringe filling rig, and a drop test apparatus with a high-speed camera - to remove manual process variability from inhaler assembly and testing, saving 6 weeks lead time against fully built devices from an external supplier.
• Ran a structured drop test failure analysis on inhaler mouthpieces. Using high-speed footage to identify crack propagation origins and latch failures, then tested ABS against PP across fully assembled devices; findings directly drove the switch from ABS to PP for the final production mouthpiece.

Six-month live design brief for BMT on an automatic day-shape display system for unmanned surface vessels. Absorbed both project manager and design lead roles across a five-person team, originally six. The project achieved a First.

• Managed all weekly client meetings with BMT, kept five engineers from different disciplines aligned across individual and group deliverables, and organised an early-stage site visit to Devon. Interviewing BMT design engineers, maintenance engineers, and commercial fishermen to establish the operational constraints that drove the system specification.
• Introduced a Google Design Sprint framework. Functional decomposition, morphological charting, and Pugh matrix selection gave the team an evidence-based path to concept convergence, replacing assumptions with justified, traceable decisions.
• Validated the design through FEA, actuator sizing, tolerance and thermal expansion analysis, and Granta Edupack material selection, confirming the sub-system met COLREGs requirements.
• Presented the finished system to 200+ engineering companies at a live design exhibition. Four of the five team members received job offers from ASV Global (now L3Harris Defence), myself included, with interest reconfirmed at a subsequent careers fair.