Antonis Sidiropoulos
About Me
Robotics Research & AI engineer with PhD in Robotics and 7+ years of experience in classical and learning-based control, motion planning and manipulation. Hands-on experience with action chunking transformers, Vision-Language-Action models, and large-scale vision models for robotic manipulation. First-author publications in top-tier robotics conferences (ICRA, IROS, Humanoids) and journals (AURO, JINT). Track record of deploying learning-based policies on industrial and agricultural robotic platforms, combining deep learning with classical control methods for safe, robust real-world performance. Passionate about embodied AI, and AI-driven autonomy for next-generation robotics.
Core Expertise
Robot Learning
- Imitation learning (BC, DAgger, Diffusion Policy, ACT)
- Reinforcement learning (PPO, SAC, DDPG)
- Vision-Language-Action models (OpenVLA, SmolVLA, pi0.5)
- Foundation models (SAM, DINO, CLIP, SigLIP)
- Perception (instance segmentation, 6D pose estimation)
Planning & Control
- Motion planning for multi-DoF manipulators (sampling & optimization based)
- Impedance/admittance control
- LQR, MPC
- Collision avoidance
Technical Stack
- PyTorch, LeRobot, HuggingFace
- ROS2, MoveIt2, ros2_control
- Isaac Sim, Mujoco, Gazebo
- C++, Python, Docker, Git
Professional Experience
AI Engineer
Oct 2025 – PresentFine-tuning and integration of LLMs and VLMs for automation and decision-making. Developing agentic workflows (ReAct, planning, tool calling) for multimodal AI applications.
Senior Robotics Software Engineer
Aug 2023 – Aug 2025Led end-to-end development of ROS 2–based motion planning and manipulation stack for industrial applications. Fine-tuned and deployed action chunking transformers. Developed learning-based motion generation combining imitation learning with classical planning. Integrated transformer-based vision models for instance segmentation and grasp planning.
Research Assistant – Robotics & AI
Dec 2016 – Dec 2022Developed imitation learning and reinforcement learning policies for agricultural manipulation, human-robot handover in assistive robotics, and collaborative manipulation. Authored multiple first-author publications in ICRA, IROS, Humanoids, AURO, and JINT. Contributed to 4 EU H2020 projects (Bacchus, RAMCIP, Collaborate, SMARTsurg).
Education
PhD in Robotics
Developed novel Learning from Demonstrations frameworks for motion generation and manipulation in dynamic environments, combining deep neural networks with dynamical systems for robust generalization. (Thesis)
B.Sc./M.Sc. in Electrical & Computer Engineering
Selected Publications
- Dynamic Via-points and Improved Spatial Generalization for Online Trajectory Generation with Dynamic Movement Primitives (2024) Journal of Intelligent & Robotic Systems (JINT)
- From rgb images to dynamic movement primitives for planar tasks (2023) IEEE-RAS Humanoids
- A novel framework for generalizing dynamic movement primitives under kinematic constraints (2023) Autonomous Robots (AURO)
- A reversible dynamic movement primitive formulation (2021) IEEE ICRA
- Human-robot collaborative object transfer using human motion prediction (2021) IEEE ICRA
- A human inspired handover policy using gaussian mixture models and haptic cues (2019) Autonomous Robots (AURO)