Recent forest fire in California creates huge amount of dust as well as pollution to the atmosphere. Conventional man-power and fighting equipment, on the other hand, are inefficient in dealing with large area forest fire. On the other hand, rainmaking rocket, which is based on existing rocket technology, can be especially effective in extinguishing a large area of forest fire. The rocket can be launched from anywhere, or even airborne, and will be detonated above the target area, releasing chemicals which accelerate rainfall process in an artificial way.
What it does
GAIA System is a next-generation, voice-controlled GCS (ground control software) based on x86 platform. It provides a solution for countering wild fire and forest protection. An unmanned spaceplane is remotely controlled by the GCS software, and the detected objects as well as the GPS coordinates of the vehicle will be rendered in real time. If a forest fire is observed, the user will be notified and he/she can select a boundary for rainmaking coverage by mouse clicking. After the user confirms, the GCS will send a command to a spaceplane to perform aerobatic maneuvers and point its nose towards the target. A high-speed rainmaking rocket can be launched towards the target to help increase rainfall rapidly, extinguishing wildfire in an artificial way. Before launch, the user can also ask about the number of GPS satellites, which will directly influence the rocket accuracy, as well as the target weather condition (in case of a rainfall, no artificial rainmaking is necessary) by voice queries.
How I built it
I started prototyping the C# application by dragging buttons around and adjusting its layout. Meanwhile, I have also spent a huge amount of time exploring vision recognition algorithms such as YOLO as well as tuning my custom REST API. For the client side application, three components are developed separately: map visualization and plane rendering, FPV panel and radar panel. Each of them takes me a significant amount of time. The client side is majorly written in C# and the server side is in Python.
Challenges I ran into
The space plane's autopilot and navigation algorithms, as well as its complex targeting mechanism take quite a while for me to figure it out correctly. Besides coding and debugging, I have also spent many hours calculating on my notebook. Besides, despite the fact that I have worked on multiple autonomous systems, I do not have access to a space plane, as well as its sensor and camera information. As a result, I have to use a "virtual space plane", which is a fairly large class file and uses my Logitech webcam as its primary camera.
Accomplishments that I'm proud of
I am able to surpass existing GCS software (such as Mission Planner and QGroundControl) by introducing many new features, including an intelligent voice agent!
What I learned
Always thinking out of the box is helpful. Despite the fact that many may regard Windows desktop applications as lackluster and old-schooled, it can be equally effective and impactful in certain scenarios. Also, the key to achieve something is by being relentless towards the end goal, and constantly learning new stuff as well as improving.
What's next for GAIA System
versions running on other platforms (MacOS, Linux or iOS/Android) are currently under development. Future versions could also utilize state-of-the-art AR/VR technology for 3D visualization of map and spaceplane. Cloud-based quantum computing such as IBM Qiskit could also be used for generating pseudo-random simulation of multiple space planes, providing better coverage for the entire country while minimizing CPU load.
Fire neutralization, forest protection as well as STEM education
Backup Video Link
(in case Youtube removes my video due to background music) https://www.youtube.com/watch?v=Rm5lvPqZa4I&feature=youtu.be