Repurposing Old Hardware - Robotic Autonomous Cameraperson
If this is your first time here, and would like to help contribute please see our Onboarding Guide. It explains the basis of the project, where it started, where it is now, and where to continue.
| Name | Role |
|---|---|
| Aleck Hernandez | Hardware Lead |
| Bevan Neiberg | AI / Computer Vision Lead |
| Briggs Tucker | Communications & UI Lead |
| Christine Morgado | Scrum Master |
| Jacob Odle | Controls Lead |
| Zoe Rizzo | Scribe |
Coach: Samuel Malachowsky
Project Methodology: Link
Project Metrics: Link
Project Final Plan: Link
Domain Model: Link
Agendas, Meeting Notes & 4-Ups: Link
RIT Software Engineering Department
Robotics, Computer Vision, Embedded Software, Raspberry Pi, Open Source, ESP32
This research project repurposes legacy robotic hardware, the ScorBot ER-4pc and ER-V, into an autonomous camera platform. The project's goal is to automate camera operations which typically require a person: panning, tilting, and framing subjects. By utilizing two robotic arms, two camera feeds can be integrated and combined, autonomously and dynamically deciding which feed to highlight. This project includes ideating, research, implementation, and refinement of software that combines machine learning, computer vision, and other technologies to guide camera movements. The previous two iterations of this project focused on subject tracking and a new controller for the ER-4pc. The plan for this year includes refining subject tracking and implementing predictive movements, a digital twin to use for testing and debugging, and an updated controller user interface for more intuitive design. This year’s main goal is multi-camera integration.
