This semester, I started to work on my senior design project. I will be working on this project for two semesters in total. Actually I is not a correct word to use for this post, a better one would be “We”. We are a team of 6. Our team consists of Sam Crane, Ladd Jones, Antonia Deliyianni and two other GT students.
Our solution aims to make search and rescue operations more efficient. We will have a swarm of autonomous drones equipped with thermal and day cameras. The drones will have companion computers with GPUs. With the on board computer, the drone will process the visual data coming from the cameras to see whether a human is spotted (or some other object of interest). If it finds something interesting, it will message the operator with images and location. The operator will look at the data and decide what to do next. The nice thing about this is, an operator would be able to use multiple drones without being have to watch all the cameras. The drones would be looking for the human, and the operator would just be dealing with other stuff.
So what we are doing right now?
Customer discovery and research, we are contacting all the possible search and rescue operator to get more information about their missions and challenges. We try to understand which conditions do they operate under and what they do to find someone. So far we’ve contacted multiple government and non-government entities and we are collecting information.
Project Design, we are outlining the technical implementation of the project and it is huge…
- The drones won’t be slaves so that is another complexity added to the problem. However it is also going to be more robust!
- There are multiple configurations for different mission locations, since we will be flying in rural areas we wont have the luxury of cellular networks. Therefore we will need to create our own network. We are planning to implement a mesh network. With that our network range can grow larger as we have more drones in the air. I already found the hardware to use for.. but it is a little pricey (300$ per unit). I wish I had more knowledge about radio comms.
- The server and the ground station implementations will also differ according to where we execute the missions. Since we will be flying in rural areas, we can’t use (lovely) AWS stuff (no internet!) so the server must be able to operate on dedicated hardware near the ground control station
- Front End UX & UI design is a thing.. We are also working on that.
- We are also designing the drone from 3D Printed materials only, that takes a good amount of time…
Looking for funding!
Thermal cameras are pricey! 1500$ therefore we would be happy if we don’t have to spend our pocket money for that.
Each drone is around 400$-500$
We have field devices, routers and ground stations that also cost a little.
Right now we are thinking about IEEE fund.