Real-Time Drone Monitoring Dashboards
Roles
Rocco Ghielmini - Software Developer
Aug, 2017 - Nov, 2017
Technologies Used
Streaming ROS telemetry to the browser
Rapyuta’s operators needed a way to watch autonomous drone fleets in real time. I built dashboards that convert high-frequency ROS topics into intuitive mission views so pilots can inspect state, sensor payloads, and alerts from anywhere.
Problem
Before this work, flight coordinators juggled separate tools for mission planning, execution, and analysis. Visualising ROS streams required desktop tooling, and sharing mission context with clients meant stitching together screenshots after the fact.
My Contributions
- Developed Preact/React dashboards that subscribe to ROS topics and render telemetry, mission state, and map overlays with minimal latency.
- Built web interfaces for designing, scheduling, and remotely piloting missions so teams could hand off routes and adjust parameters mid-flight.
- Implemented Redux modules that keep mission timelines in sync across operators and automatically capture audit trails.
- Prototyped Python-based image classification pipelines to flag anomalies, replacing a manual review loop with automated alerts.
Outcomes
Operators can now control and monitor drone fleets from a single browser session while clients follow along live. Automated classification reduced the time engineers spent labeling footage, letting the team scale deployments without increasing headcount.