Selected work / My contribution & the results

Robotics & software.

Physical robots, the software that controls them, and tools that make engineering work easier. Explore my contribution, the design decisions and what exists today.

Coursework: T.H.O.M.A.S.Company programs: S.L.A.B., Raven & VIER

7 case studies

Machines & the systems behind them

Team-built platforms: from physical structure to control, perception and experiments.

Hexapod · Aetheron Labs / team project

Novexa

An 18-joint hexapod built from a modular, 3D-printed structure, servo-driven legs and an Arduino-based control system.

My contribution: Mechanical design, fabrication & integration
What you can inspect: Physical prototype · CAD files · Journal paper

Experimental prototypeRead the case study

Robot control · Collaborative platform

OSRS

A robotics platform that connects hardware, camera feedback, simulation and a live control dashboard for repeatable experiments.

My contribution: Platform engineering & online system identification
What you can inspect: Source code · 20-trial SIL results · Technical report

Open-source research platformRead the case study

Software for specific problems

Research, work and decision tools, with each project’s purpose and development stage kept clear.

Actual RoboScout static torque calculator showing inputs, calculated torque and engineering assumptions

Actual application · Static torque calculator

Engineering research · Open source

RoboScout

An engineering research tool that connects component selection to source documents, torque calculations and reviewable reports.

My contribution: Product direction & engineering workflow design
What you can inspect: Application source · Architecture · Sizing routines

Public software projectRead the case study

Actual WIRK local workspace showing the starter task dashboard and focus controls

Local-only preview · Bundled starter data

Productivity · Community edition

WIRK

A browser-based workspace that keeps task planning, focus sessions and a private journal together, with an optional AI assistant.

My contribution: Independent community-edition project
What you can inspect: Application source · Local storage · Import / export

Open-source community editionRead the case study

Actual RecoverOS command center running its built-in simulation with synthetic payment data, not real revenue

Simulation / test mode · Not real revenue

Decision systems · Prototype

RecoverOS

A payment-recovery prototype that checks proposed retry actions against policy rules and records outcomes only after gateway confirmation.

My contribution: Independent decision-system prototype
What you can inspect: Source code · Synthetic evaluations · Test-mode flow

Simulation and test-mode prototypeRead the case study

Actual Calm Pulse web prototype welcome screen with its static sample conversation and entry form

Web prototype · Static sample conversation

Wellness app · Team prototype

Calm Pulse

A mobile wellness prototype combining mood and sleep tracking, journaling, breathing exercises and AI-assisted conversation.

My contribution: Back-end development within a four-person team
What you can inspect: React Native app source · Credited team roles

Team-built prototypeRead the case study

Actual Nook local document preview showing an audit-created page that explicitly labels simulated capabilities

Experimental workspace · Review-created sample data

AI workspace · Experimental prototype

Nook

An open-source workspace prototype connecting editable documents, coworker profiles and model-provider interfaces, with simulated computer tools.

My contribution: Independent repository · Workspace prototype
What you can inspect: Working document editor · Local persistence · Inspectable route handlers

Experimental local workspaceRead the case study

Coursework / Robotics / Since May 2026

T.H.O.M.A.S.
Learning to find a route.

An autonomous maze-solving robot developed as AI & Robotics coursework. The work starts in simulation and is moving toward physical hardware.

Navigation without a map

The simulated robot uses Q-learning and line-sensor observations to learn navigation without access to a map. My CV records training and testing across different maze layouts in MuJoCo.

The next stage covers mechanical design, electronics and control-system integration. No physical maze-solving result or measured success rate is claimed here.

Discuss robot learning

Context: AI & Robotics coursework
Current stage: Simulation work; physical-hardware development
Technical focus: Q-learning · MuJoCo · Line-sensor navigation
Available record: Project description in my CV · No separate demo or test report is linked yet.

Earlier work & smaller repositories

More from the archive.

Smaller applications and programming exercises. Browse the code on GitHub; the selected work above offers a closer look at my current interests.

video-downloader-apiPython python-mini-projectsPython NumberGamePython TrueRecipesJavaScript Book-SantaJavaScript Explore Aetheron Labs and its programs