Our company develops bespoke software applications, algorithms, and machine learning solutions for clients in the US, Europe and Asia. We specialise in remote condition monitoring and railway signalling applications, however we have also developed applications for fintech and commercial services providers where our experience with large datasets can be leveraged.We are experts in the following:Railway signalling equipment - We have a decade of experience working data from all types of wayside signalling equipment, including track switches, track circuits, relay interlockings, CBTC systems, OLE and more. We know their failure modes, how they are maintained, and the engineering processes associated with them, both in the UK and overseas. Python and the machine learning stack - We have developed all kinds of applications in Python, from proofs-of-concept using Tensorflow and Pandas, to full fledged customer facing web applications running on AWS serving thousands of users.

- Led the development and deployment of condition monitoring algorithms and provided reliability and failure analysis for trackside and signaling equipment for a specialist supplier of railway condition monitoring and datalogging equipment. - Developed reliability analysis and diagnostics algorithms in Python and Java by analyzing historical data and FMS and Ellipse records.- Designed and implemented machine learning algorithms in Python and Java, using Support Vector Machines, Bayesian Classification, GLM, and Deep Learning with Tensorflow.- Consulted individual Network Rail engineers as well as overseas infrastructure operators on degraded mode operation signaling and track switch fault prognostics.- Reproduced trackside equipment and railway switch failures, and studied their effects by performing experiments at manufacturer sites and test facilities. - Developed a condition monitoring algorithim for Network Rail's electric and hydraulic points machines that improved failure prediction rates from 10% to 25%. - Developed an alarm threshold automation algorithm for track circuits and points machines which reduced customer RCM software deployment time from hours to minutes. - Developed a track circuit performance monitoring algorithm which allowed customers to automatically monitor the performance of a variety of track circuit devices.- Developed a image recognition algorithms for level crossing monitoring using CNNs in Tensorflow.- Carried out business development activities, and led successful bids for datalogging project contracts for both metro and mainline applications.- Secured partnerships with two major multinational systems integrators, offering consultancy, software and hardware.- Organised reliability centered maintenance (RCM) tool training events for clients.- Delivered technical and change management consultancy to senior Network Rail managers.

- Designed and built a prototype railway track switch with passive fault tolerance capabilities by researching the reliability and safety of railway track switches as part of a three-member team.- Researched and analyzed railway track switch failures by analyzing Network Rail FMS records, studying related research, and building dynamic UK track switch mathematic models.- Constructed electrical control systems and software for a lab prototype track switch using C++, Matlab, Kollmorgen Servodrive, and motor hardware.- Designed a custom three-phase supply cabinet to power prototype rig.- Performed specialist motion control and failure analysis using dSpace hardware and software.- Awarded three UK and overseas track switching technology patents.- Secured three years follow up research funding worth over £2MM.- Secured the involvement of major multinational railway technology suppliers which included providing test sites, funding, and engineering team support.- Produced technical reports, wiring diagrams, CAD drawings, and technical documentation for the project.- Presented the project at academic conferences and industry events in the UK and overseas.- Instructed six Masters and PhD students on railway S&C, electrical engineering, and control systems by providing lab-based motion control and hands-on track switching technology instruction.- Discussed the electrical, control, and fault tolerance aspects of the project in daily engineering meetings.- Gave technical presentations and held quarterly review meetings with sponsors and industrial partners.

Control Systems Research- Conducted PhD research in control systems and condition monitoring.- Wrote academic reports and peer reviewed research, presented at UKACC 2012 and SOFT 2012.University Work- Administered three hour lab sessions for second year students studying control system design, and to final year students studying sensors and actuators.- Invigilated exams for students with additional needs.

Control Systems Research - Designed a model based condition monitoring scheme for the Joint European Torus (JET) cryogenic vacuum system, deployed on the neutral beam heating systems. This is the first scheme of its type.- Carried out a comprehensive FMECA and statistical failure analysis for a cryogenic vacuum system.- Used Kalman filtering and residual evaluation to provide fault detection and isolation (FDI) capability to the scheme, which was tested and validated using historical process data.-Sponsored by the Culham Centre for Fusion Energy (CCFE), as part of an industrial CASE award.- Published peer reviewed research in academic journals and conferences, including Fusion Engineering and Design, and UKACC.- Delivered technical presentations and reports to engineering staff, academics and management.