Responsibilities include:

• Designed and developed an integrated task and motion planning framework for executing real-world cooking tasks using a dual-arm robotic system.
• Integrated the PDDLStream library with MoveIt Task Constructor multi-stage manipulation planner to enhance multi-step motion planning for interdependent tasks.
• Formulated PDDL-based predicates and action templates based on our proposed dual-arm slicing strategy and implemented policies to handle uncertainty in a dynamic environment, such as perception-based replanning.
• Demonstrated the proposed framework in simulation and real robot experiments through a case study of a cucumber slicing task. Augmented our framework with various cooking-related skills, such as object fixturing, force-based tip detection, and slicing using Reinforcement Learning (RL).
• Preliminary results of the proposed framework were accepted as a contribution (paper and demo video) and will be presented as a poster in the Cooking Robotics Workshop at ICRA 2024.

Responsibilities include:

• Proposed and developed a novel bin-picking system using vision and 3D tactile sensors with an approved R&D budget of 1.5 million yen (Approx. $10k USD).
• Developed custom Python algorithms for performing adaptive grasping and slip detection from tactile sensor data.
• Developed custom Python-based middleware for external communication between TM5-900 collaborative robot, 2F-85 Robotiq gripper, and 3D tactile sensors using socket communication protocol.
• Received official industrial robot operator training & certification from Omron Automation, covering the required safety standards, legal regulations, and general knowledge related to industrial robots in Japan.