Tamer Guven | Autonomous Systems, Robotics and Computer Vision

Projects

Engineering across software and hardware.

Selected work in autonomous vehicles, competition robotics, machine learning, research software and full-stack systems.

01 In Development Personal engineering project

ROS 2 Autonomous Vehicle Platform

An ongoing personal project to design and build an autonomous 1:8-scale RC vehicle from the mechanical platform to the control software. The system uses ROS 2 Jazzy with a modular simulation-to-hardware architecture for perception, planning, control and hardware interfaces.

My contribution

Designing the vehicle platform, ROS 2 package structure, URDF/Xacro model, compatible simulation and hardware interfaces, and the planned steering and throttle control stack.

ROS 2 JazzyGazeboRVizURDF/Xacroros2_controlPythonRaspberry PiComputer VisionIMUEmbedded ControlSolidWorks

Project details GitHub repository

Simulation-to-hardware architecture

Gazebo simulation Physical vehicle

Compatible ROS 2 interfaces

Sensors and Perception

State Estimation

Planning

Vehicle Control

ros2_control and Hardware Interface

SteeringThrottle

Active development. The repository currently focuses on robot description and mock control setup.

Read full technical description

Objective

Develop a 1:8-scale autonomous RC vehicle with a modular architecture that can move from Gazebo simulation to physical hardware through compatible ROS 2 interfaces.

Current progress

The public repository currently contains a Xacro vehicle model, RViz configuration, display launch file, bicycle steering controller configuration and a mock ros2_control setup. A permanent hardware interface has not yet been added.

Mechanical and embedded plan

The platform will initially use a brushless DC motor. After mechanical validation, a two-cylinder four-stroke nitro engine may be integrated. A Raspberry Pi 4, camera, IMU, steering servo and throttle servo are planned for onboard control and perception.

Development approach

Perception, planning, control, simulation and hardware interfaces are separated. Early integration can use fabricated CSV or JSON perception inputs before the camera pipeline is connected.

Status is intentionally presented as ongoing. The repository states that the current mock setup is not ready to control an unrestricted physical vehicle.

02 Competition robotics Team engineering

FRC Autonomous, Vision and Control Systems

Developed and mentored Java and WPILib software for competition robots, including autonomous routines, sensor-based control, AprilTag vision, odometry and real-time robot control. Worked across software, electrical and mechanical teams to build reliable systems for competition environments.

My contribution

Developed autonomous and vision systems, built multi-camera AprilTag localisation, supported PID-based control and path recovery, led software strategy, and trained students across Java, CAN wiring, PWM, encoders and vision integration.

JavaWPILibLimelightAprilTagSwerve DrivePID ControlOdometryComputer VisionCAN BusGit

Project details Mostra code Flare code

Robot localisation and control

Limelight cameras Wheel encoders Gyroscope

Robot Pose Estimation

Autonomous and Driver-Controlled Logic

SwerveMechanismsTargeting

AprilTag pose estimates are evaluated and combined with encoder and gyro odometry before use by robot logic.

Read full technical description

Vision and localisation

Multiple Limelight cameras produce AprilTag-based robot pose estimates. These measurements are evaluated before being integrated with encoder and gyro odometry to improve field localisation.

Autonomous control

Created predefined autonomous trajectories using odometry and PID correction so the robot could detect deviations and recover from collisions, game pieces and other disturbances.

Mechanism calculations

Worked on odometry-based moving-shot turret calculations, target tracking and required shooting vectors. This work is presented as competition development, not as a completed commercial targeting system.

Cross-team integration

Coordinated software with electrical and mechanical work through Git and hands-on debugging. Training covered Java, CAN, PWM, power distribution, encoders, beam-break sensors and vision integration.

View original project details

Archived wording retained verbatim. The concise case study above uses only claims supported by the current CV and public repositories.

FRC Autonomous Navigation and Control System

Designed and implemented the full autonomous control stack for a 120-pound competitive robot in the FIRST Robotics Competition. Developed full simulation of robot systems to validate control algorithms before deployment. The system executed pre-planned trajectories with real-time odometry correction, achieving consistent sub-inch positioning accuracy across a 54-by-27-foot field.

Technical Challenge

The robot needed to autonomously score game pieces during a 15-second period with no driver input. This required precise motion profiling, encoder-based dead reckoning, and fault-tolerant state machine logic to handle sensor noise and mechanical slop in real time.

Architecture

Built on the WPILib command-based framework. Implemented PID controllers for drivetrain velocity and heading, a trapezoidal motion profiler for smooth acceleration, and a finite state machine for sequencing autonomous actions. Integrated gyroscope and wheel encoder fusion for continuous pose estimation.

Innovation in Control Award-winning autonomous control system at regional competition

03 In Development Open-source project

OpenHotel

An open-source, local-first hotel management system for small and medium hotels, with booking, guest, check-in, billing, housekeeping and operational workflows.

My contribution

Built the full-stack project, designed REST APIs and role-based access control, and established foundations for secure ESP32 keycard devices and room access logging.

Java 21Spring BootSpring SecurityReactTypeScriptMySQLDockerREST APIs

Project details GitHub repository

Local-first platform architecture

React + TypeScript client

Spring Boot REST API

MySQLAccess logs

Future ESP32 keycard devices

Docker Compose runs the local frontend, backend, database and database viewer with matching versions.

Read full technical description

Objective

Provide a local-first hotel platform for booking and day-to-day operations while keeping local data and uploaded documents out of the public source repository.

Current scope

The repository documents guest accounts, room booking, check-in and check-out, billing, activities, housekeeping, manager tools, ID document review and room access logs.

Security foundations

The project uses Spring Security, role-based access control and configurable environment values. ESP32 keycard integration is described as a future device capability.

Status

The CV dates the project from July 2026 to present. The repository is public, but no live production deployment is claimed.

04 Completed research project Machine learning

Second-Hand Car Price Prediction

Developed a Python machine-learning project using historical vehicle data to predict second-hand car prices. The work included data analysis, model development, evaluation and an authored research paper documenting the methodology and results.

My contribution

Cleaned and prepared the Germany Used Cars Dataset 2023, encoded categorical features, built and evaluated a neural-network regression model, and authored the accompanying research paper.

Dataset200,000+ offers

Model10-layer neural network

MetricMean squared error

Final loss0.3216 train · 0.2513 validation

PythonpandasNumPyscikit-learnTensorFlow/KerasJupyter NotebookKagglePredictive Modelling

Project details Kaggle notebook Research paper

Regression workflow

Historical vehicle data

Cleaning + validation

One-hot encoding + scaling

Neural network regression

Price predictionMSE evaluation

The paper concludes that supervised learning can provide a price estimate, while noting that predictions are not exact.

Read full technical description

Problem

Explore whether supervised learning can estimate second-hand car prices from historical listing data without presenting the result as financial advice.

Data preparation

Validated inconsistent categorical values and registration-date formats, removed empty rows, selected relevant features, applied one-hot encoding and split the data into training and validation sets.

Model and evaluation

The paper documents a 10-layer neural network trained with SGD, a learning rate of 0.00001, momentum of 0.8 and mean squared error. The reported final losses are shown above.

Conclusion

The paper concludes that supervised learning can predict an indicative second-hand price, while acknowledging that the estimates are not exact.

View original project details

Archived wording retained verbatim. The current paper describes a 10-layer model, so that verified value is used in the summary above.

Machine Learning: Second-Hand Car Price Prediction

Built a complete machine learning system to predict second-hand car prices from the Germany Used Cars dataset. Implemented data cleaning, feature engineering, categorical encoding, and trained a deep neural network regression model. The system handles 8+ car features and predicts prices with optimized loss convergence.

Technical Challenge

The dataset contained inconsistent categorical values, mixed data types, missing entries, and non-standard formats in critical columns (registration dates, numeric fields). Required custom validation logic to detect and remove invalid records while preserving dataset size. Categorical features needed encoding while maintaining model generalization.

Architecture

Implemented a 9-layer sequential neural network with 100 units per dense layer. Applied one-hot encoding for categorical features (model, color, transmission type, fuel type). Used StandardScaler for feature and target normalization. Trained with SGD optimizer (learning rate: 0.00001, momentum: 0.8) with weight decay for regularization. Tracked training and validation loss over 30 epochs to monitor convergence and prevent overfitting.

View on Kaggle

05 Research Project Bogazici University

Rodent Behaviour Analysis

Developed Python software at Bogazici University to analyse rodent behavioural patterns using machine learning and image processing in controlled experimental scenarios.

My contribution

Built Python computer-vision and machine-learning software, processed behavioural observations from recordings, and collaborated with researchers and senior students as the youngest member of the team.

PythonMachine LearningOpenCVImage ProcessingComputer VisionBehaviour Analysis

Project details

Research analysis workflow

Controlled experiment recordings

Image processing

Machine learning analysis

Behavioural observations

Research prototype. No public project repository is linked in the current portfolio or CV.

Read full technical description

Objective

Analyse recorded rodent behaviour in controlled experimental scenarios using Python, computer vision, image processing and machine learning.

Research contribution

Processed and interpreted behavioural observations while working alongside older and more experienced students as the youngest member of the research team.

Status and availability

The work is presented as a university research project. No public code or deployment claim is made.

View original project details

Archived wording retained verbatim. The concise summary above does not repeat the original worldwide-use or open-source claims because no matching public repository is linked.

Analysis of Rodent Behavior Patterns with Machine Learning

Developed computer vision and machine learning system to track rodent movement and identify behavioral patterns across different environments. Built free software tool designed for laboratory use worldwide, enabling researchers to analyze animal behavior autonomously and consistently across diverse experimental conditions.

Technical Challenge

The primary challenge was maintaining robust tracking across diverse environments and different rodent morphologies. Environmental variations (lighting conditions, cage setups, background textures) and rodent-specific factors (size, fur color, movement speed) required adaptive vision algorithms that could generalize without requiring manual tuning for each experiment.

Architecture

Implemented computer vision system using OpenCV for real-time rodent detection and tracking. Applied background subtraction, morphological operations, and contour analysis to isolate rodents from complex environments. Integrated machine learning classifiers to identify behavioral patterns from trajectory data. Developed configurable software interface allowing researchers to adapt detection parameters for different rodent species and experimental setups without programming knowledge.

Open-source behavioral analysis tool for research labs

Experience

Robotics leadership, research and software development.

Roles are presented with student and team context, with original portfolio descriptions retained for reference.

Robotics Software and Electrical Mentor

FIRST Robotics Team Mostra, Istanbul Technical University

Nov 2024 – Apr 2026

Mentored students across Java, WPILib, electrical systems, autonomous control and multi-camera AprilTag localisation while coordinating software strategy with the wider robot team.

Built a Java and WPILib multi-camera AprilTag localisation system that combined evaluated vision estimates with encoder and gyro odometry.
Created odometry-based autonomous trajectories with PID correction and recovery from disturbances.
Trained more than 15 students in competition-ready electrical systems, including CAN, PWM, power distribution, sensors, encoders and vision integration.
Led through teaching, review and debugging so students could understand and develop the systems themselves.

GitHub Repos

More about this experience

Original portfolio wording retained below. The title and dates above follow the current CV.

Scaled Java software training program across 15+ engineers; designed curriculum covering real-time systems, sensor fusion, and vision processing
Architected autonomous control systems with low-latency feedback loops and adaptive vision integration; optimized for 120-lb industrial robot platform
Led software strategy for Innovation in Control Award-winning autonomous and control systems

Team Captain and Software Lead

FIRST Robotics Team Flare, FMV Erenkoy Isik High School

Sep 2021 – Jun 2025

Software Team Leader · Jun 2022 to Sep 2024 Team Captain · Jul 2023 to Sep 2024

Progressed from software and mechanical team member to leading a 40-student team, coordinating three subteams and contributing to three competition robots.

Developed Java and WPILib autonomous, vision and robot-control systems.
Led and trained the software subteam while coordinating software, mechanical and electrical integration through Git.
Led SolidWorks design and supported manufacturing and assembly for competition robots.
Contributed to robots that received competition recognition, including a finalist result and autonomous systems award.

GitHub Repos

More about this experience

Original portfolio wording retained below. The title and 40-student team size above follow the current CV.

Led team strategy and managed 3 subteams (50+ engineers) to deliver tournament finalist robot; coordinated mechanical, electrical, and software integration
Built autonomous control and vision systems in Java/WPILib for 3 competitive robots; one achieved award-winning autonomous performance under hard real-time constraints
Architected real-time sensor fusion systems and state machine logic; established Git-based engineering workflows and code review standards across organization
Mentored and trained software subteam; designed educational curriculum covering robotics fundamentals, PID control, and vision algorithms
Designed and built 3 production robots in SolidWorks; one became tournament finalist, demonstrating mechanical precision and engineering rigor

Machine Learning Researcher

Georgia Institute of Technology research project

Jun 2023 – Sep 2023

Developed a used-car price prediction project in Python using historical vehicle data, conducted data analysis and predictive modelling, and authored a research paper documenting the methodology and results.

View Research Paper View Kaggle Notebook

More about this experience

Original portfolio wording retained below. The paper is described above as authored, not formally published.

Georgia Institute of Technology

Engineered end-to-end ML system for regression modeling on structured automotive dataset; published research paper documenting methodology and results
Developed predictive models using Python (TensorFlow, scikit-learn); optimized feature engineering and hyperparameter tuning for model convergence
Implemented unsupervised learning techniques for exploratory data analysis; gained experience with production-grade ML workflows and statistical validation

Software Developer Intern

Token Financial Technologies

Aug 2023 – Sep 2023

Used GitLab for version control and development workflows, learned professional software development processes, collaborated in a cross-functional team and supported project delivery.

More about this experience

Original portfolio wording retained below.

Contributed to production backend systems within cross-functional fintech engineering team
Adopted GitLab-based version control workflows; participated in code review cycles and CI/CD deployment processes
Collaborated with senior engineers on scalable system design; gained exposure to production software development practices and team coordination

Research Assistant

Bogazici University

Sep 2021 – May 2022

Developed Python software using machine learning and image processing to analyse rodent behaviour in controlled experimental scenarios, working alongside older and more experienced students as the youngest member of the team.

More about this experience

Original portfolio wording retained below.

Developed Python-based computer vision system for behavioral analysis; processed video data using OpenCV and machine learning for rodent tracking and classification
Analyzed and interpreted animal behavioral patterns in controlled experimental scenarios; contributed to reproducible experimental workflows
As freshman member, collaborated with senior researchers on data-driven behavioral science; gained experience in academic research methodologies and statistical analysis

Awards

Awards and recognition.

External recognition for robotics engineering, autonomous control, team leadership and technical impact.

Year not listed

Innovation in Control Award

FRC Team Mostra · Regional competition

Recognised the team's autonomous and control systems during Tamer's software and electrical mentoring period.

Year not listed

Autonomous Systems Award

FRC Team Flare · Competition result

Recognised the precise autonomous performance of a competition robot developed with Java, WPILib and odometry-based control.

Year not listed

Tournament Finalist

FRC Team Flare · Competition result

Team finalist result achieved after coordinated software, mechanical and electrical development across the robot programme.

The available site and CV do not identify exact event names, award years or official result links, so none are inferred here.

Technical Skills

Tools connected to real project work.

Grouped by engineering area without percentages or self-assigned proficiency scores.

Robotics and Autonomous Systems

ROS 2GazeboRVizWPILibAprilTag VisionOdometryPID ControlTrajectory PlanningSensor IntegrationSwerve Driveros2_controlState Machine DesignReal-Time Systems

ROS 2 → Autonomous Vehicle Platform · WPILib → FRC Autonomous Systems

Programming

JavaPythonCC++SQLHTMLCSSRISC-V AssemblyMU0 AssemblySystemVerilogMySQL

Java → FRC and OpenHotel · Python → ML and research software

Computer Vision and Machine Learning

OpenCVTensorFlowKerasscikit-learnpandasNumPyImage ProcessingPose EstimationJupyter NotebookKaggleMachine Learning

scikit-learn → Car Price Prediction · OpenCV → Rodent Behaviour Analysis

Engineering and Development Tools

GitGitHubGitLabLinux / CLIDockerVS CodeIntelliJ IDEAJupyter NotebooksSolidWorksRaspberry PiObject-Oriented DesignCommand-Based ArchitectureUnit TestingSystems EngineeringControl TheoryAutonomous Navigation

SolidWorks → Competition robots and RC vehicle · Docker → OpenHotel

About

Software that connects with the physical world.

I am a Computer Science student at The University of Manchester interested in autonomous vehicles, robotics and intelligent systems. My experience spans robot software, computer vision, autonomous control, electrical integration, mechanical design and machine learning. Through FIRST Robotics, university research and personal engineering projects, I have learned to build systems that connect software with real hardware and to work effectively across technical teams.

I am particularly interested in autonomous systems, embedded software, perception, controls and vehicle technology, and in continuing to develop as an engineer.

More about my background

Earlier portfolio wording is retained verbatim below. The recruiter-facing summary above uses the current CV and project evidence.

Autonomous Systems Engineer

I build autonomous systems and real-time control software that operate under hard constraints. I lead engineering teams, architect complex system solutions, and focus on deterministic, production-grade performance. Passionate about robotics and systems engineering.

I approach engineering from the system level down. Whether it is an autonomous robot navigating a field under time constraints or a perception system classifying objects at 30 fps, I focus on architecture that is robust, deterministic, and measurable.

My foundation is in competitive robotics, where I led the design and implementation of autonomous routines, sensor fusion logic, and real-time control loops for 120-pound industrial robots under the FIRST Robotics Competition framework. These systems operated under hard timing constraints with no margin for software failure.

I have extended this systems thinking into machine learning research at Georgia Tech and Bogazici University, and into production software development at Token Financial Technologies. I am driven by the intersection of perception, control, and decision-making in physical systems.

Tamer Guven — 2026

Contact

Get in Touch

For questions about my projects, research or robotics work, email me or connect with me on LinkedIn and GitHub.

Email Me Download CV LinkedIn GitHub