HealthDrones - navegação de VANTs autônomos baseada em autômatos celulares

Abstract

Instituted by Law 9985/2000, the National System of Conservation Units regulates management and oversight of natural resources in environmental preservation areas. However, the large amount of Conservation Units in different locations with a wide range of territorial extension weakens its oversight regarding enviromental crimes. As such, using techonologies capable of assisting in the process of inspection can be of great help. In this work we investigate ways to assist this process by means of a detailed analysis of autonomous unmanned aerial vehicles (UAVs). This work’s objective is to define and validate a computational model capable of performing manouvers and flights to inspect a certain area. The possibility of using these UAVs in an autonomous and controlled manner was analyzed. Experiments were performed to measure automated flight manouvers accuracy, and the results obtained were compatible with what the manufacturer atests to deliver when it comes to the remote control software of the UAV. The same maneuvers were tested using the maker’s remote control software and the automated scripts developed using a Node.js API, with mostly identical average results, with a confidence interval of 95%. Besides, a simulation environment was implemented along with a cellular automaton capable of generating pathways and executing the vehicle’s simulated actions. The scripts generated by the automaton follow the same developing patterns of the scripts in the first experiments, showing that it is statistically compatible with the maker’s software. Finally, solutions to make the studied UAV more autonomous for environmental preservation areas inspection are proposed, and problems with selected UAV are pointed.