Sensor Fusion for ArduPilot Academic Research

Hey all. I am looking for any combination of assistance or colaboration in a project of mine I am planning to start. I am an Aerospace Engineering graduate student who is starting to plan out my project for academic research (Special project -> PhD Thesis). My interest involves using systems engineering algorithms to further advance the use of sensor fusion in UAVs for various missions such as object avoidance through urban areas (GPS unavailable/low-light/manueverability).

The idea is that in situations where the UAV is in difficult situations (GPS-denied, reduced light, difficult objects for lidars, high manueverability needed) then you must take advantage of ALL sensors on-board, including the IMU, and you must also have a complex algorithm which has a built-up knowledge of the surroundings.

My proposal is to begin to work in both the physical world and the simulation world. The physical world involves a Pixhawk and some basic sensors. The simulation world would be a built up modeling and simulation environment to work with the ArduPilot code. It will utilize complex algorithms involving UAV dynamics, sensor communication, and machine learning.

On the programming and software side I will need a lot of help. I do not have an advanced background in programming other than engineering programming on like Matlab and with Numpy in Python. BUT this is why I want to start the project NOW, and get assistance in learning what I need to learn. I am willing to research as hard as I need to in order to get the knowledge.

If anyone has detailed questions you can reach me at: harris.caleb84@gmail.com. Otherwise, just comment here.

1 Like

G’day Caleb. I’m not able to help you, but suggest swinging by https://gitter.im/ArduPilot/Research and letting that channel know what you are trying to do.

I am heading over there now! Thanks James!