José Martínez Carranza of the National Institute of Astrophysics, Optics, and Electronics (INAOE) has developed a new navigation system for autonomous drones that does not rely on GPS. Normally, autonomous drones rely on GPS to figure out where they are and to navigate to where they need to be. Unfortunately, GPS isn’t perfect, and there are spots where data isn’t available or where wind currents or other environmental factors make it hard to pinpoint a location or keep data updated.
Carranza’s new system replaces GPS with low-cost accelerometers, gyroscopes, and camcorders that allow the drones to scan their environment for visual clues, figure out where they are, and determine how to proceed. The drone still needs to know where it’s going in order to use visual data to navigate to its destination, but this is true of the GPS models as well. Generally, these drones are programmed with a specific path to follow and points at which to stop and collect data. The new system simply allows them to do that like humans and other animals have done for millions of years: without GPS satellites.
The project began in the United Kingdom, where Carranza did post-doctoral at the University of Bristol. He worked with Blue Bear, Ltd., which produced the drones for the experimentation. Now that he’s back in Mexico, his work will continue to improve the system and develop better ways for autonomous drones to move about. He thinks that such drones can have a wide range of uses, as they already do, but would like to see them used to address a wider variety of issues than just military use or package delivery. Surveillance is one, allowing police to keep an eye on things in more dynamic ways that simply using CCTV, but using them to explore towers and tall buildings for structural problems that would be impossible to ascertain from inside might be another.