Peter Burke, IT specialist Peter Burke Artificial intelligence can autonomously generate control systems – “Brains” – other robots, performing complex coding tasks much faster than traditional human teams. The project uses advanced AI generative models, including ChatgPT, Gemini and Claude, to create a fully functional drone control system that works completely on board the drone.
Burke, a professor of electrical engineering and computer science at the University of California in Irvine, structured the project around two types of “robots”. The first is AI software operating on laptops and in the cloud, responsible for writing the code. The second is the drone itself, which uses Raspberry Pi Zero 2 W for hosting and starting software generated by AI in real time.
Traditional drone systems are based on ground control software, such as planner Mission or Qustoncontrol for flight management. The Burke approach replaces the WebgCS (Web Ground Control Station) system controlling the control station, which operates directly on the drone. This allows you to access live desktop via a standard web browser, ensuring telemetry in real time, mission planning and autonomous navigation.
The development process was organized in four intensive sprints. The first sprint used Claude in the browser to generate the initial code database, but memory limits prevented the completion of the project. Later attempts from Gemini 2.5 and Cursor IDE improved functionality, but they encountered errors such as problems with bash shell scripts and limitations of context in many files.
The fourth and last sprint, using Windsurf IDE, allowed AI to effectively produce the WebGCS system. Over 2.5 weeks and about 100 hours of human work, AI generated 10,000 code lines, including Python scripts, HTML, JavaScript and Bash. It is about 20 times faster than the previous Burke project, Cloudstation, which required four years of cumulative work of the student team.
In the project, he emphasized current limitations in AI coding: while models can effectively support codes databases to about 10,000 lines, the performance is rapidly lowered for larger systems. Studies confirm that exceeding the limits of tokens in AI models leads to a reduction in the accuracy of code generation and debugging.
The implications of this work go beyond drones. By showing that artificial intelligence can autonomously create complex, multilingual software systems, the Burke project indicates the future in which machines can design and manage other machines. Although the current system remains limited to individual drones, research suggests the potential of swarms controlled by AI, autonomous applications of spatial intelligence and high -scale automated control systems.
Such technologies can radically transform the field of aviation robotics, thanks to which autonomous navigation, planning and decision making is more available. However, questions about reliability, testing in unpredictable environments and security supervision remain the main challenges for the future of robotics based on AI.
