-Led a team of engineers and researchers to develop robotic applications like bin picking, piece sorting, welding, and polishing, applying state-of-the-art research to deliver practical vision-based robotics solutions. -Architected and built the back-end infrastructure for vision-based robotic systems using Flask and SQL databases to manage data and organize input/output. Collaborated with the UI team to ensure seamless integration between front-end and back-end components. - Managed Continuous Integration (CI) and version control with GitHub, implemented GitHub Actions for automated builds and testing. Utilized Docker for containerization, ensuring consistent environments across development, testing, and deployment stages. Directed branching strategies, code reviews, and merging processes, while overseeing manual unit testing to maintain code quality and facilitate team collaboration.

-Experienced with 3D sensors like stereo vision, Time-of-Flight (TOF), LiDAR, and structured light, enabling reliable object perception in various environments. Proficient in deploying these systems on edge computing platforms for real-time, efficient processing. -Implemented advanced object detection models such as YOLO, Mask R-CNN, and DETR using PyTorch to accurately detect and classify objects in complex environments. -Developed and implemented 6D pose estimation solutions by integrating PoseCNN, PoseBERT, and ICP techniques for accurate estimation and fine-tuning of object poses. Utilized 3D data structures, such as octrees and kd-trees, to efficiently manage spatial data and accelerate related processing tasks within the pose estimation pipeline. -Implemented and thoroughly tested OMPL's path planning algorithms for complex, continuous paths with collision avoidance using ROS/MoveIt. -Conducted research on path planning for industrial robots using NVIDIA Isaac Sim as a simulation environment to test and validate Deep Reinforcement Learning (DRL) techniques, such as Proximal Policy Optimization (PPO) and Soft Actor-Critic (SAC). Developed advanced path planning strategies with a specific focus on challenging applications like robotic welding.

Developed detection algorithms for vision-based robotic pick-and-place solution for solar panels.