• Program and optimize 6‑axis industrial manipulators (FANUC, NACHI, KAWASAKI) for multi‑variant car body assembly—including high‑

precision roof line adaptations—using TP programming, I/O mapping, and DCS safety zones to ensure synchronized tool trajectories in fast‑

paced production environments.

• Deploy robot programs directly on the production floor, conducting real‑time tuning and Sim2Real calibration to reduce cycle time, minimize

idle movements, and ensure robust execution under line conditions with varying payloads and tolerances.

• Rigorously validate robot motion programs through teach pendant simulation, path optimization, and real‑time collision checks, prioritizing

technician safety and customer reliability in accordance with Toyota’s strict safety protocols and JIS/ISO standards.

• Interface with cross‑functional teams—SEBI (Safety Engineering for Body Integration), Digital Engineering, Equipment Design, and Quality

Assurance—to cross‑verify process feasibility, interlock logic, and fail‑safe sequences across PLCs and robot controllers.

• Utilize proprietary Toyota toolchains and data visualization platforms (such as G2, RSLogix, and safety matrix validators) to monitor execution

flows, mitigate risks through fault‑tree analysis, and maintain interdepartmental traceability for zero‑defect manufacturing compliance.

• Deployed NVIDIA Isaac ROS Visual Odometry and SLAM on an Nvidia Jetson Orin Nano‑powered Robot Dog (along with simulation) to perform motion tracking, pose estimation and 3D reconstruction using a Realsense D455 RGBD camera. Simulated the robot behavior on Isaac Sim, and visualized the SLAM‑related data like VO path, SLAM path, Odometry, Velocity, images and 3D point clouds on RViz and rqt.

• Developed a robust localization stack for the robot by integrating data from 2D RPLidar, IMU, and stereo cameras. Implemented AMCL‑based localization on pre‑loaded maps, and enabled real‑time SLAM using only Lidar data in the absence of wheel odometry. Utilized Luxnois camera boards for processing and stereo depth sensing.

• Conducted Sensor fusion of RP‑LiDAR, RGBD Camera, Laser sensor, and IMU for robust Navigation stack; performed extensive debugging to resolve sensor discrepancies and ensure coherent multi‑sensor data fusion.

• Collected 3D point cloud data, which was later converted into 2D occupancy grid to develop a .PGM map. Conducted Semantic Segmentation for accurate feature labeling and validation of map quality prior to web deployment.

• Established inter‑system integration between ROS2 and ROS1 (via Docker containers) for real‑time teleoperation; debugged network and node‑bridge issues for seamless cross‑platform control from ROS2.0 to the robot via the node in Docker.

• Designed and developed the mechanical architecture of the autonomous robot dog (humanoid‑legged layout) using Onshape and SolidWorks; conducted validation through FEA tools like Hyperworks. Prototyped mechanical fixtures for the robot, finalized designs for vendor manufacturing, and resolved mechanical, CAN, and Ethernet issues to ensure reliable autonomous operation.

• Assisted in the testing and analysis of various materials for a specific project, evaluating their compatibility with project requirements. Conducted tensile, compression, and hardness tests for material selection and future design decisions

• Supported the engineering team in creating detailed CAD drawings for components and assemblies. Contributed to the documentation of design changes, ensuring accurate and up-to-date records. Developed proficiency in AutoCAD and contributed to maintaining an organized and accessible design database

• Participated in the prototyping phase of a new product, assisting in the fabrication and assembly of prototypes. Contributed to the testing process, documenting results and providing feedback for design improvements

• Applied Random Forest and Gradient Boosting ML models to predict potential equipment failures for a proposed robotic assembly line. Integrated these models with a priority-based scheduling algorithm to plan maintenance during non-peak hours to boost efficiency by reducing the overall downtime by 29%