Palaniappan Yeagappan

About

Robotics software engineer focused on real-time state estimation and perception pipelines, with hands-on C++/ROS work improving LiDAR+IMU odometry stability via an iterative EKF and careful timing/synchronization fixes. Has integrated LIO-SAM, built multi-robot communication bridges (ROS + custom UDP with heartbeat/fallback), and uses Gazebo + Docker for repeatable testing, backed by CI/CD experience maintaining Azure DevOps pipelines at Cognizant.

Experience

Education

Key Strengths

• Deep hands-on state estimation and sensor fusion (LiDAR/IMU) for autonomous navigation
• Implemented and integrated an iterative EKF in C++ within a ROS pipeline (callbacks, measurement models, interfaces)
• Strong real-time debugging methodology using log replay, timestamp alignment checks, and profiling
• Improved odometry reliability by fixing sync offsets and tuning IMU noise parameters to reduce drift
• Built robust heterogeneous robot communication with standardized interfaces, UDP low-latency path, and heartbeat/fallback logic
• Practical simulation-first validation in Gazebo for TF/timing and behavior before hardware runs
• DevOps strength: maintained Azure DevOps CI/CD pipelines for automated build/test/deploy
• Deployed real-time vision object detection on a live conveyor line integrated with a PLC
• Systematic latency/throughput debugging using a replayable test harness and end-to-end timestamping
• Improved model robustness by collecting on-line data and applying targeted augmentation plus ROI tightening
• Practical production-floor fixes (diffused lighting/hood, vibration mounting) to stabilize vision performance
• Cross-domain debugging across software, hardware, and networking; identified unreliable switch port under load
• Implemented reliability mechanisms (sequence numbers, retry logic) to eliminate missed PLC commands (missed-sort rate to 0)
• Built customer-specific Python reporting (real-time alerts, trend analysis, weekly summaries) integrated via JSON/config
• On-site collaboration with operators/engineers; applied Lean/5S to improve assembly workflow efficiency (~1.59%)
• Hands-on IIoT stack development with ESP32 sensor nodes and cloud analytics