Project Overview

Typically, when a fleet of drones is employed for various surveillance purposes and/or search-and-rescue missions over a given area, one of the crucial aspects is to ensure that the response-time is as bounded as possible for every point within that geo-area. However, in many practical scenarios, there are obstacles (a.k.a. "no-fly") zones due to various constraints (e.g., tall buildings in urban settings; proximity of airports; military installations; protected eco-zones; etc.) -- which need to be taken into account when considering the needed flight-time to ensure observational coverage.

The main goal of this project is to implement algorithms for partitioning a given geo-area into zones with the awareness of no-fly polygonal obstacles, in a manner that for a given fleet of drones, their distribution will ensure a minimal average response time in each partition for which a specific drone may be responsible for monitoring. The algorithms are part of the current research by the client/advisor and his team, and will be provided for the project. The objective is to develop a prototype system that will enable visualization of the partitioning (for a given geo-region; set of obstacles; and set of drones) and show a sample of a drone's flight for a given location-of-interest and the starting position of the drone). The system should be designed to have a UI that will enable the users to see the configuration (based on the info in back-end files) and allow the users to select the source for that configuration (i.e., a specific file), as well as interactively select a location in the final partition-scheme to show the response of the drone in charge for that zone.