Robotics Software Engineer

Ömer Mersin

SLAM & Perception for Autonomous Systems

Robotics software engineer building real-time localization, perception, and guidance pipelines for autonomous aerial and ground platforms. Currently at Dronetools in Sevilla, Spain.

Robotics Software Engineer SLAM Engineer Perception Engineer
C++ Python ROS 2 Sensor Fusion Real-time Systems LiDAR-IMU Fusion
Open to Robotics Roles
Netherlands (NL) Switzerland (CH) Germany (DE)
Explore My Work View Resume Get in Touch

About Me

Ömer Mersin

2+ Years

Robotics Software Engineering

I design and ship autonomy software for robots operating in GPS-denied and dynamic environments. My focus is SLAM, state estimation, navigation, and perception pipelines that must run reliably in real time.

At Dronetools, I work on autonomous drone navigation and operator systems that connect low-level control, sensor fusion, and mission workflows. I also competed in TEKNOFEST Robotaxi for 3 consecutive years with Team Mekatek, improving the system every year and helping the team reach 3rd place plus Best Team Spirit Award. I am open to relocation for robotics roles in the Netherlands, Switzerland, and Germany.

Robotics Skill Matrix

Core Robotics Engineering

C++ Python ROS 2 Real-time Systems Multithreaded Pipelines Docker

Perception & SLAM

LiDAR-Inertial Odometry FAST-LIO Sensor Fusion OpenCV YOLO-based Detection Video Stabilization & Tracking

Navigation, Control & Deployment

3D ESDF Planning Autonomous Decision-Making MAVLink (PX4/ArduPilot) GStreamer/RTSP CUDA + NVIDIA Jetson WPF/Qt Operator Tools
Flagship Work

SLAM & Autonomy Portfolio

Four flagship and award-winning systems that demonstrate end-to-end robotics capability: perception, real-time mapping, planning, controls, and deployable software architecture.

LIONav SLAM mapping output GIF
Flagship Project

LIONav

Modular ROS 2 autonomy stack for GPS-denied drone navigation, combining FAST-LIO odometry, GPU ESDF mapping, pose-graph consistency, multi-level planning, Smart RTL, safety failsafes, MAVLink integration, and real-time GCS bridges.

Launch orchestration of 11 nodes + 5 included launches + web telemetry server
Smart RTL with hybrid dynamic tail + static corridor memory and continuous ESDF replanning
6-level autonomy manager with automatic planner fallback plus independent safety watchdog
Project manager for multi-stream recording/replay and field mission lifecycle tooling

System Scope

11 core packages, ~43k LOC across core pipeline

Interfaces

40 custom ROS interfaces (planner/project managers)

Ground Control

WebSocket bridge, UDP native bridge, Qt6 native GCS

ROS 2 Humble FAST-LIO nvblox TSDF/ESDF GTSAM iSAM2 Smart RTL Planner Manager Safety Watchdog MAVLink Web + Native GCS Bridges
View Source
Smart Gimbal Manager Interface
Flagship Project

Smart Gimbal Manager

Professional-grade UAV ground control station (GCS) for real-time gimbal, payload, telemetry, video, and mapping workflows across surveillance, inspection, and emergency operations.

Multi-protocol gimbal control (BaseCam, Viewlink, DJI) with clean abstraction layers
Dual map/video operations with camera footprint projection and terrain intersection tools
Low-latency telemetry + video pipeline with recording, replay, and geospatial export workflows

Platform

Windows WPF desktop app on .NET 6 (C#)

Architecture

MVVM + modular solution structure (40+ projects)

Ops Scope

Gimbal, mapping, video, cloud telemetry, and mission tooling

C# .NET 6 WPF MVVM GMap.NET ArcGIS/Cesium MQTT FFmpeg + OpenCV
View Source
Flagship Project

Video Stabilization Engine

Real-time, GPU-accelerated video stabilization and ultra-low-latency RTSP streaming library with seamless passthrough or processing modes, optimized for NVIDIA Jetson edge deployments.

Dual-mode RTSP: zero-processing passthrough or full processing with hot switching
GPU stabilization, roll correction (Canny + Hough), and configurable smoothing modes
DeepStream detection/tracking and Jetson hardware-encoder path for minimum latency

Latency

~10-20 ms passthrough, ~50-100 ms processing mode

Architecture

gstd + gst-interpipe with instant mode switching

Robotics Use

Stabilized FPV/perception streams for drones and robots

C++17 OpenCV CUDA GStreamer + gstd gst-interpipe DeepStream RTSP NVIDIA Jetson
View Source
TEKNOFEST Robotaxi Team Mekatek
Award-Winning Project

TEKNOFEST Robotaxi (Team Mekatek)

Autonomous vehicle competition project where our team achieved 3rd place and the Best Team Spirit Award. I contributed to real-time perception and sensor-fusion decision making under track constraints.

Drivable-area detection using image-processing pipelines
Label/sign detection with YOLO for real-time scene understanding
Sensor-fusion based autonomous driving and decision logic
Competed 3 years in a row and improved system robustness each season

Awards

3rd Place + Best Team Spirit Award

Perception

Drivable area + YOLO label detection

Autonomy

Multi-sensor fusion for driving decisions

Autonomous Driving Image Processing YOLO Sensor Fusion Real-time Decision Making
Team Proof Competition

Additional robotics work and experiments are available on GitHub.

Connect

Let's Work Together

Open to SLAM, robotics perception, and autonomous navigation opportunities across Europe.

Available for NL / CH / DE roles

I am actively exploring robotics software roles focused on autonomy, localization, and navigation in the Netherlands, Switzerland, and Germany. Open to relocation and remote collaboration in European time zones.

GitHub

Check out my repositories

LinkedIn

Let's connect professionally