Embedded and robotics
Microcontroller code, sensor integration, autonomous behaviour, and hardware-aware debugging.
Cambridge Engineering student
Usairam Abdullah
I build and debug technical systems across embedded control, robotics, real-time graphics, simulation, machine learning, and self-hosted infrastructure.
- Current focus
- Systems, controls, graphics, and applied ML
- Preferred work
- Projects with hardware, simulation, or deployed software
- Base
- Birmingham / Cambridge
Technical focus
Engineering work with clear constraints
My projects sit close to the boundary between software and physical systems: sensor data, control loops, rendering pipelines, training data, deployment choices, and the debugging work needed to make those pieces behave.
Graphics and simulation
OpenGL, transformations, terrain, physics, camera control, and visual tools for understanding system state.
Infrastructure and ML
Static hosting, private networking, service management, CNN training workflows, evaluation, and documented limitations.
Experience
Research and engineering in practice
Alongside independent projects, I have worked on research and team engineering problems where the value is in turning sensing, analysis, and physical systems into something useful and testable.
Research / 3D data / Spatial analysis
Point-Cloud Research Experience
Research experience focused on working with point-cloud data: understanding three-dimensional scenes, exploring how spatial information can be represented, and communicating findings through a visual technical workflow.
Scroll or use the controls to view the capture and reconstruction stages.
4SB / Robotics / Embedded systems
4SB AGV Development
Engineering work on an autonomous guided vehicle, bringing together sensing, embedded control, movement behaviour, and practical testing of a physical robotics system.
The work connected software decisions with the realities of hardware integration, repeatable motion, and on-device debugging.
Featured projects
Selected technical work
These projects are written as engineering records: what was built, which constraints mattered, and where the useful debugging happened.
C++ / OpenGL / Control systems
Mars Lander Simulation
A simulated lander environment covering vehicle dynamics, terrain-aware behaviour, landing and crash logic, and visual debugging tools for understanding simulation state.
- Modelled descent behaviour and collision outcomes.
- Used generated terrain to test landing constraints.
- Built visual feedback for debugging motion and state.
C++ / OpenGL / GLSL
OpenGL Real-Time Rendering Engine
A graphics project focused on the fundamentals of a rendering pipeline: shaders, transformations, camera control, linear algebra, and frame-by-frame inspection.
- Implemented shader-driven rendering behaviour.
- Worked with transformations and camera movement.
- Used graphics fundamentals to reason about scene output.
MicroPython / Raspberry Pi Pico / Sensors
Miniature Delivery AGV
A small autonomous vehicle project combining embedded control, sensor input, movement logic, and practical hardware/software debugging on a constrained microcontroller platform.
- Integrated sensor readings into vehicle behaviour.
- Developed MicroPython control logic on Raspberry Pi Pico.
- Debugged issues across code, wiring, and physical motion.
Ubuntu Server / Docker / Tailscale / Cloudflare
Self-hosted Home Server / Portfolio Infrastructure
A home server setup treated as a maintainable deployment project: static hosting, service isolation, remote access, public/private separation, and routine security hygiene.
- Organised services around clear access boundaries.
- Used Docker/CasaOS-style workflows for service management.
- Prepared static portfolio hosting for an Ubuntu server.
Python / TensorFlow / Dataset labelling
CNN Image Classification
An image classification workflow built around a self-labelled dataset, training pipeline, model evaluation, and careful notes on limitations and next improvements.
- Prepared labelled image data for supervised training.
- Trained and evaluated a TensorFlow CNN model.
- Documented likely failure modes and improvement routes.
Skills and toolkit
Tools grouped by the work they support
Programming
C++, Python, MicroPython, JavaScript, HTML, CSS
Graphics and simulation
OpenGL, GLSL, linear algebra, camera systems, physics logic
Embedded systems
Raspberry Pi Pico, sensors, control logic, hardware debugging
Machine learning
TensorFlow, CNNs, dataset labelling, training, evaluation
Infrastructure
Ubuntu Server, Docker, CasaOS, Tailscale, Cloudflare, Nginx
Working style
Debugging, documentation, iterative testing, readable handoff notes
Self-hosted infrastructure
A deployment project, not just a place to put files
The home server setup is included as an engineering project because it has real operational constraints: keeping static hosting simple, separating public services from private access, managing containers, and documenting how the site can be updated without relying on a complex build pipeline.
Infrastructure notes
- Ubuntu Server as the base deployment environment.
- Docker or CasaOS for service management.
- Tailscale for private access where appropriate.
- Cloudflare Tunnel or Nginx for public static hosting.
- No secrets, keys, private IPs, or internal hostnames in this site.
About
Technical systems, written clearly
I am a Cambridge Engineering student interested in systems that connect computation with real constraints: embedded hardware, robotics, graphics, simulation, control, machine learning, and infrastructure. I value projects where the evidence is visible in the implementation, the tradeoffs are documented, and the next debugging step is clear.
CV and education
Engineering education
I am completing an MEng in Engineering at the University of Cambridge and entering my third year. My studies cover electronics, mechanics, linear systems, structures, electrical engineering, thermodynamics, materials, and mathematics. Before university, I achieved three A*s in Mathematics, Further Mathematics, and Physics, alongside an A in Chemistry and an award-winning EPQ on AI in healthcare.
Contact
For internships, project work, and technical collaboration
The best contact route is email. GitHub and LinkedIn links can be added once the public profiles are ready.