Robotics • Controls • Embedded

Engineer building reliable autonomy pipelines — from sensors to control.

ROS2-first stack across GNSS/NTRIP, wheel‑speed/IMU fusion (EKF), audio siren detection, CAN telemetry, and containerized deployments (Docker, CI/CD). Shipping to real hardware like NXP S32G & Raspberry Pi.

Realtime ROS2 nodes: 15+ u‑blox F9P + SPARTN Audio ML (TFLite)

Highlighted projects

PointPerfect NTRIP Client (ROS2)

Resilient TLS client that auto‑discovers mountpoints, sends live GGA, streams SPARTN/RTCM3 corrections to a u‑blox receiver. Minimal CPU, auto‑reconnect, and metrics heartbeat. Also includes real‑time GNSS correction pipeline and slip‑aware EKF refinements for stability.

PythonTLSSPARTNUBXRTCM3 ROS2 Humbleu‑blox F9PDocker

Engine Control & Optimization

Supported engine start and idle control profiling using full feedback LQR with fast‑path actuation to minimize delays and ensure closed loop torque. Developed tactical features including cost‑based optimization (predictive control) enhancements and physics‑based equations deployment.

ControlsOptimizationPredictive Control

Vx/Vy EKF with Slip‑Aware Fusion

Nine‑state EKF fusing wheel speeds, IMU yaw‑rate/accel, GPS velocities. Dynamic noise scaling under slip, low‑speed yaw freeze, and delayed measurement handling via buffer replay.

MATLAB/SimulinkEKFSensor Fusion

Emergency Vehicle Detection (Audio ML)

Binary siren detector derived from YAMNet embeddings with a custom TFLite model. Real‑time ROS2 node with amplitude topic, ducking logic, and CPU‑efficient resampling.

TFLiteAudio DSPROS2

FleetPlayback (Alerts & Adverts)

Priority‑based audio mixer/player with ALSA + libsndfile. JSON‑driven mapping, duck/restore envelopes, and concurrent playback with ROS2 QoS tuning.

C++ALSAJSON

CAN Telemetry + Web Dashboard

SocketCAN receivers publishing to ROS2, with a lightweight web dashboard for monitoring GPS fix, wheel speeds, IMU, and system state. Dockerized with CI/CD to edge hardware.

SocketCANDockerCI/CDWeb

Experience

Education & Highlights

  • Controls & Mechatronics background; deep MATLAB/Simulink experience.
  • Certified in Machine Learning with practical applications.
  • 3 years of professional Python development experience.
  • Specialized in torque path controls including cost optimization, shift energy management, creep, launch control, e‑coasting, and brake blending.

Skills

Core

ROS2PythonC++ MATLAB/SimulinkDocker CI/CDLinux

Perception & Fusion

EKFMPCIMU GPS/GNSSNTRIPpyubx2

Platforms & I/O

NXP S32GRaspberry Pi ALSASocketCANI²C/UART

Résumé

Want the full details? Download a concise résumé PDF.

Download PDF

Contact