Slide 1 Slide 2 Slide 3

Kulunu Samarawickrama

Doctoral Researcher, Tampere University, Finland

Introduction

I am a doctoral researcher in the Cognitive Robotics Research Group at Tampere University, Finland. My work focuses on advancing 3D perception techniques to empower robotic systems with capabilities like grasping and assembly. I am driven by a desire to bridge the gap between cutting-edge research and its practical applications in industries. My fascination with how machines learn and adapt to their environments stems from years of academic exploration and hands-on experience.

My Journey

The Mechatronics Engineer

My academic journey began with a bachelor's degree in mechatronics at the Asian Institute of Technology, Thailand. During this time, I cultivated a strong foundation in mechanics, electronics, and programming. Post-graduation, I delved into the automation industry, working on challenging projects involving industrial robots and vision-based inspection systems. These experiences laid the groundwork for my growing passion for robotics and AI, inspiring me to explore further studies.

The Researcher and Developer

My master’s journey at Tampere University allowed me to specialize in robotics and artificial intelligence. Immersed in courses like computer vision and machine learning, I honed my ability to design and implement intelligent systems. My thesis, focused on 3D perception for robotic manipulation, earned accolades for its innovation and rigor. Today, I channel this expertise into collaborative projects aimed at integrating robotics and AI into real-world applications.

Research Focus

My PhD research is at the intersection of cognitive science and robotics, where I work on building models that enable robots to think and act intelligently. My focus areas include:

  • Enhancing 3D visual perception for improved object detection and pose estimation
  • Developing advanced algorithms for point cloud segmentation and analysis
  • Creating synthetic datasets to simulate real-world scenarios for robotic training
  • Integrating perception models into robotic systems via ROS for seamless automation
  • Establishing robust benchmarks to evaluate model performance and accuracy

Publications

  • Automatic Dataset Generation From CAD for Vision-Based Grasping
    Ahmad, Samarawickrama, K., Rahtu, E., & Pieters, R. (2021).
    2021 20th International Conference on Advanced Robotics (ICAR), 715–721.
    Read more
  • Sensor-based human–robot collaboration for industrial tasks
    Alexandre Angleraud, Akif Ekrekli, Kulunu Samarawickrama, Gaurang Sharma, Roel Pieters
    Robotics and Computer-Integrated Manufacturing, Volume 86, 2024.
    Read more