Samuel Appleby

About

Samuel Appleby
Programmer

Phone: +44 7465439846

I am a computer science postgraduate who has just passed my PhD thesis defence Newcastle University. I have been programming since the start of my bachelor's degree in September 2017. My preferred languages are: Python, Kotlin, C#, C++, Java and C. In my own time, I enjoy developing my own projects and systems. This includes my Pokémon analyser, which has also been used by other developers in the industry of Pokémon game balancing. Most recently, I have developed an Android phone application for sports markets analysis. Further details can be found in the projects section. My GitHub can be found here.

Education

My most recent academic experience is passing my thesis defence (02/02/2026) for my PhD Computer Science. My thesis is titled:

On the Modelling of Temporal Data
Linear Logics in Event Logs and Deep Reinforcement Learning in Real-Time Simulations

Which focuses on bridging the gap between temporal modelling as found in Business Process Management (BPM) versus those in real-time systems exploited by (deep) reinforcement learning. Other major milestones include:

PhD Computer Science (2022–2026) Newcastle University

MSc Computer Game Engineering (2020–2021) Distinction Newcastle University

BSc Computer Science (2017–2020) 1:1 Newcastle University

Recent Experience

Prior to my PhD and during my MSc in Computer Game Engineering, I was a junior network programmer at Lucid Games Ltd. developing Destruction AllStars, a new release for the (also newly released) PlayStation 5. Here, I became familiar with REST API technologies and Node.js. My significant contributions include:

Managing message protocols between the client and the web service, specifically, the integration of filtering into their matchmaking system.
Extending Unreal's Gauntlet's Automation Framework. Here, I re-adapted the automation tests to communicate and synchronise with the web service, allowing for greater control from the network. I also wrote the HTTP request system for this project.

During my time at Lucid Games Ltd, I acquired extensive networking knowledge which I was able to transfer across to my own personal projects.

Recent Projects

My most recent project is called Arbiter, an abritrage betting application for Android. Combining my work and academic experience, I was able to produce a pipeline with many components leveraging the following:

Databases technologies and querying, including PostgreSQL and SQLite
Cloud platform hosting from Microsoft Azure
Cloud messaging software using Google Firebase Cloud Messaging
UI design using Android Jetpack Compose
UX design for an Android application

I have future plans for this project and hope to publish it on the Google Play Store once copyright and licencing is acquired.

Projects

Personal

I have a few personal projects which I have worked on in my own time, located on my GitHub here.

Arbiter

Android Application. Arbitrage engine for sports markets. Covers 3 continents, including 10+ countries. Over 60 bookmakers. Multiple markets. Cloud-hosted on Microsoft Azure. Live Firebase notifications.

Probability Theory Statistics Market Analysis Risk Management

Kotlin Jetpack Compose Python PostgreSQL SQLite Azure Firebase API Querying REST API

Pokémon Team Analyser

A Pokémon type analyser, providing the optimal team combinations per configuration. Uses Google API to communicate with online spreadsheet datasets. Used for a community-developed Pokémon spin-off game.

Data Analysis Statistics

C# Google Sheet API

PhD Computer Science

One major element of my PhD was the creation of a deep reinforcement learning pipeline. This involved the creation of a Unity Simulation and learning pipeline. Several state-of-the-art deep reinforcement learning algorithms were used to train an agent for active target tracking.

Soft Actor-Critic (SAC)

Proximal Policy Optimisation (PPO)

Twin Delayed Deep Deterministic Policy Gradients (TD3)

Soft Actor-Critic (SAC): Noisy

MSc Computer Game Engineering

Team Project

The team project during my MSc degree involved the recreation of a Fall Guys knock-off. Our team built a custom game engine from a skeleton codebase using a range of middleware. The following showcase the levels from my team project (Feb-Mar, 2021).

Bounce Guys

Monke Kong

Cannon Hell

Sandbox

Alpha Build

Individual Work

During my BSc and MSc, modules involved the development of game technologies. The following demonstrate those.

Game Technology Innovations

Jan-Feb, 2021. Using a commercial game engine to prototype a "Fall Guys" knock-off. Available to play here!

Advanced Game Technologies

Dec, 2020-Jan, 2021. In a homebrew C++ game engine, prototype a "Fall Guys" knock-off.

Advanced Graphics For Games

Nov-Dec, 2020. Create a Cyberpunk setting using OpenGL.

BSc Computer Science

Computer Games Development

Jan-May, 2020. Create a video game using any engine/technologies. Available to play here!

Dissertation

Sep, 2019 - May, 2020. Final year undergraduate dissertation project using Goal-Oriented Action Planning for game AI.

Gaming Simulations

Jan-May, 2020. Finish the physics engine, input and game logic for an arcade game.

Papers

Articles

Appleby, S., Ushaw, G. (2025). From Camera Image to Active Target Tracking: Modelling, Encoding and Metrical Analysis for Unmanned Underwater Vehicles. AI 2025, 6(4), 71. 10.3390/ai6040071

Deep Learning Reinforcement Learning Active Target Tracking Computer Vision Sim-to-Real Unity Simulation

Bergami, G., Appleby, S., Morgan, G. (2023). Specification Mining over Temporal Data. Computers 2023, 12(9), 185. 10.3390/computers12090185

Specification Mining Temporal Data Algorithm Optimisation Heuristic Search

Bergami, G., Appleby, S. (2023). Quickening Data-Aware Conformance Checking through Temporal Algebras. Information 2023, 14(3), 173. 10.3390/info14030173

Conformance Checking Temporal Algebras Business Process Management

Conference Proceedings

Appleby, S., Crane, K., Bergami, G., McGough, A. S. (2025). SWiMM DEEPeR: A Simulated Underwater Environment for Tracking Marine Mammals Using Deep Reinforcement Learning and BlueROV2. AI 2025, 6(4), 71. 10.1109/CoG57401.2023.10333168