ADS-B Drone Collision Avoidance System

— Project Description —

The project being proposed by Group 18 in conjunction with Nevada Dynamics (ND) is to take the web based drone ground control system ND has in place and integrate a collision avoidance module for interaction with aircraft in the surrounding area that transmit ADS-B signals.

Design
The project’s first stage is to develop a software package that can interface with an antenna placed on the drones, gather the ADS-B signals, and filter out the useless or corrupted signals. The program will gather the signals coming in, filter out unusable signals, and update a JSON formatted file for parsing every few seconds.

Secondly, an algorithm would be developed with polar math, to use the position, heading, and speed of both the aircraft and the drone and detect whether a collision or near miss is going to happen. Specifically, polar math will be used on these data points to determine a circular path around the globe that each object is on at the current time. If the intersection is close enough, the drone will be stopped by using ND’s API. Once the signal gathering and collision avoidance model are finished, they will be made available for drones using Nevada Dynamics’ ground control platform.

Target Audience
The main users will be drone owners and operators that take advantage of ND’s platform, which will be most useful for survey and delivery drones flying at decently high altitude. Since the project will be integrated into the drone ground control system, that platform’s usability is the important factor, and it is being designed in a very user-friendly fashion with a well made graphical user interface as well as some abstraction in the drone controls for reduced confusion with new drone users.

Operating Platform
The project will be written in Python, on a Linux platform to start with, and possibly be written for Android in Java. Since this project is being written on an embedded system, memory and processor management will be key in the optimization of the algorithm.