Autonomous Fleet Monitoring & Remote-Control Platform

Principal Engineer · cloud architecture lead · TURING DRIVE · 2020 – 2023

AWS
MQTT
ROS
Autoware
WebSocket
Node.js
Vue
IoT Core

Autonomous Fleet Monitoring & Remote-Control Platform

A real-time monitoring and remote-control platform for an autonomous fleet, designed and built from scratch on AWS, integrating 13 autonomous vehicles across 4 types: self-driving shuttles, smart golf carts, and remotely operable vehicles.

The platform subscribes to on-vehicle ROS topics via the AWS IoT SDK and streams to the cloud over MQTT. The path is bidirectional: a user issues a control command in the browser, the system receives the MQTT signal and converts it into a ROS message for the vehicle to execute, at sub-second latency. The backend is composed of EC2 / Lambda / S3 / API Gateway, with a full-stack responsive dashboard for live status, historical analysis, and remote intervention.

I also founded and led the Cloud Operations Team that built the platform’s development workflow, ops standards, and best practices, ensuring scalability and stability.

Self-driving golf-cart fleet OCC: live fleet positions on a satellite map, rent/return status, speed and routes.

Port AV/work-vehicle fleet OCC: live positions + geofence-zone monitoring (point-in-polygon: which zone a vehicle is in / whether it left), vehicles driving the port roads.