CV
Curriculum vitae, experience, education, and research profile of Kulunu Samarawickrama.
Contact Information
| Name | Kulunu Samarawickrama |
| Professional Title | Robotics & Autonomous Systems | 3D Perception |
| kulunuds@gmail.com | |
| Location | Helsinki, Finland |
Professional Summary
I work on robot perception, with a focus on 3D computer vision, ROS2 systems, and manipulation pipelines. My background combines research and hands-on software engineering: building perception stacks, generating simulation data, running GPU experiments, and turning ideas into reproducible code that others can build on. I am especially interested in practical robotics problems, clean engineering, and software that is clear, reproducible, and useful to other researchers and developers.
Experience
-
2020 - 2026 Tampere, Finland
Doctoral Researcher
Cognitive Robotics Research Group, Tampere University
Robotics & 3D Computer Vision. Building end-to-end ROS2 perception stacks and simulation-to-real pipelines with deep learning inference for robotic manipulation.
- Built end-to-end ROS2 perception stacks integrating RGB-D sensors with deep learning inference for detection, segmentation, and 6-DoF pose estimation.
- Delivered simulation-to-real pipelines using CAD-based synthetic data to scale training/benchmarking and validate model generalization.
- Engineered automated pipelines for dataset generation, metrological evaluation, and reproducible experiments across local and multi-GPU HPC environments.
- Integrated perception modules with ROS2-based system interfaces and control components to enable task-level execution.
-
2020 - 2020 Helsinki, Finland
Test Assistant
Zen Robotics Oy
Supported testing and benchmarking workflows for large-scale industrial robotic systems; improved reliability of verification runs.
- Testing and benchmarking of industrial robotic systems
- Reliability and verification improvements
-
2017 - 2018 Phan Thong, Thailand
Production Engineer
Thai Steel Cable PCL
Integrated industrial robots, machine vision, and PLC-based automation for production systems; coordinated cross-functional troubleshooting.
- Industrial robot and machine vision integration
- PLC-based automation and production system coordination
Education
-
2022 - 2026 Tampere, Finland
-
2019 - 2021 Tampere, Finland
M.Sc.
Tampere University
Artificial Intelligence & Robotics
- Distinction
- Master’s Thesis: RGB-D Deep Learning for Robotic Perception
-
2014 - 2017 Bangkok, Thailand
Awards
-
2023 ICRA 2023 METRICS ADAPT - 1st Place
IEEE ICRA
1st place (Video Track & Live Demo) for Sim2Real pose estimation challenge
-
2022 ADAPT Field Campaign 2022 - 2nd Place
EU Horizon 2020 ADAPT Project
2nd place (Video Track) for collaborative assembly challenge
Publications
-
2024 Sensor-based human-robot collaboration for industrial tasks
Robotics and Computer-Integrated Manufacturing
Alexandre Angleraud, Akif Ekrekli, Kulunu Samarawickrama, Gaurang Sharma, and Roel Pieters. DOI: 10.1016/j.rcim.2023.102663.
-
2024 6DoF assembly pose estimation dataset for robotic manipulation
Data in Brief
Kulunu Samarawickrama and Roel Pieters. DOI: 10.1016/j.dib.2024.110834.
-
2024 6D Assembly Pose Estimation by Point Cloud Registration for Robotic Manipulation
IEEE International Conference on Automation Science and Engineering
Kulunu Samarawickrama, Gaurang Sharma, Alexandre Angleraud, and Roel Pieters.
-
2021 Automatic Dataset Generation from CAD for Vision-Based Grasping
International Conference on Advanced Robotics
Kulunu Samarawickrama and others.
Skills
Languages
Certificates
-
Deep Learning
- Aalto University
- Convolutional neural networks, recurrent neural networks, attention-based models, and graph neural networks
- Deep learning with few labeled examples, deep autoencoders, flow-based and autoregressive generative models
- Generative adversarial networks and unsupervised learning via denoising
-
Build Basic Generative Adversarial Networks (GANs)
- Coursera
- Fundamentals of GANs and their applications
- Implementation of multiple GAN architectures
Contributed Projects
-
Practical and Reliable Foundation Models for Robot Manipulation (PERFORM, RCF)
Contributed to research on practical and reliable foundation-model methods for robot manipulation, with emphasis on perception, reproducible software, and robotics system integration.
- Robot manipulation and perception workflows
- Foundation-model-oriented robotics software
- Reproducible experimentation and system integration
-
Helsinki Institute of Physics Technology Programme (ROBOT)
Contributed to robotics research and technology development through the Helsinki Institute of Physics Technology Programme project ROBOT.
- Robotics technology development
- Perception software and system integration
- Research software for robotic manipulation workflows
-
OpenDR - Open Deep Learning Toolkit for Robotics
Contributed to work under the European Union’s Horizon 2020 research and innovation programme, grant agreement no. 871449.
- Deep-learning-based robotics perception
- Open-source robotics software
- Integration of perception components into robotic systems
-
METRICS - Metrological Evaluation and Testing of Robots in International Competitions
Contributed to work under the European Union’s Horizon 2020 research and innovation programme, grant agreement no. 871252.
- Robotics benchmarking and metrological evaluation
- Simulation-to-real validation pipelines
- ICRA 2023 METRICS ADAPT - 1st place (Video Track & Live Demo)
- ADAPT Field Campaign 2022 - 2nd place (Video Track)