MULAGIRI SANDEEP

Developed Yaskawa industrial robot programs using MotoSim to automate spray-coating of manufactured hand tools. Optimized robot trajectories for different tool geometries and sizes, using camera-based position estimation to support accurate tool positioning and consistent coating coverage. Validated robot programs through simulation and production trials to ensure reliable and repeatable coating across different tool variants.

Collected and annotated 600+ images of Cars and motorcycles and performed data augmentation to create diverse training data. Converted annotations to YOLOv8 format using Python and prepared the dataset for model training. Trained and evaluated a YOLOv8 object-detection model, achieving 87% mAP on the test set and validating detection performance. Deployed the trained model using ONNX Runtime and TensorRT for optimized inference.

Software Development System Analysis & Design

Validated joining processes using FEA simulations in Abaqus, correlating experimental results with simulation outcomes for higher accuracy. Developed and optimized process parameters through iterative testing, ensuring improved material performance and joint quality. Applied simulation-driven design and analysis to predict system behavior and enhance reliability in real-world applications.