Before the mapping software hit your phone, pure experience helped you decide which route you would take to get out of your house and back. At specific times of the day, it’s best to avoid a certain crossroads. At other times, you are on a roll because you would be going against the grain. It sounds pretty straightforward. But how does a drone – an unmanned aerial vehicle – decide the fastest route for it?
Before I try to answer this question, here are a few more: Does a drone have to decide which route to take? Yes. Are not the roads just straight lines in the air, as the crow flies? Certainly not, as the research of Professor Ashwini Ratnoo shows.
Perception and planning are the two pillars on which the usefulness of a drone depends. The drone must be able to perceive where it is, what is around it, and then choose an appropriate path to its destination, says Professor Ratnoo of the Department of Aerospace Engineering at the Indian Institute of Science in Bengaluru.
Ratnoo’s lab at the institute is working on the planning part of this “brain”. For a drone to function autonomously, it must be programmed with algorithms that make human decisions with precision, says the September edition of the institute’s newsletter. Core.
“Externally, if we have, say, three drones to serve five destinations, then we have to program them to optimize the total time taken,” explains the professor.
Beyond the optimization, the code must take into account the capacity of the drone. “The trajectories of the drones must be decided. A sharp turn may not be possible. In addition, we must ensure that the algorithm does not become too computationally intensive to the point of rendering the drone inefficient. “
The inefficiency here is not just a matter of time. Ratnoo says, “If the drone is in a confined area, it can wait for the algorithm to calculate and tell it what to do. But if it is facing a dynamic object, such as another drone approaching it, the algorithm should help the drone quickly decide what to do.
Drones can be useful in a myriad of areas – warfare, agriculture, anti-poaching measures, rescue efforts in stricken non-navigable areas, and providing emergency health care, among others.
What concepts are Ratnoo and his team using to make drones more efficient and safer? “Control systems are a big part of our job. Control theory is used worldwide for the guidance and control of vehicles. Ratnoo also has a favorite that helps his team stay in control of the efficiency aspect – the fork theory. “Bifurcation theory helps us use the same control algorithm structure to vary control design parameters. So just like flipping a single switch, I can use the same algorithm to come up with various trajectories that the drone could use. Additionally, keeping the same feedback structure in the control algorithm presents the best use of the drone’s detection capabilities.
This brings us to an even more interesting part of Ratnoo’s job: When multiple drones are in action, what decides their path? His team worked on “Drone Skyways” in collaboration with the institute’s Robert Bosch Center for Cyber Physical Systems (RBCCPS). The algorithms help create a corridor, or a virtual route, as the newsletter article says, for UAVs to move around safely. Drones fly at an altitude called “class G,” which is closest to the earth’s surface, he explains.
In collaboration with his colleague, Professor Debasish Ghose, his group participated in the development of “CORRIDRONE”, a gateway for the movement of several drones which also includes a geo-fence, that is to say a virtual fence along from the hall. And, these hallways can be dynamic. “Which means that once a drone passes a location, the geo-fence around that space disappears and another drone can use it freely.”
“If two drones are in different lanes but likely to cross, the algorithm helps decide whether one should accelerate and clear the junction quickly or wait and let the other drone pass,” says Ratnoo.
Does this mean that every drone carries with it the equivalent of an airport’s air traffic control system that helps prevent mid-air collisions? Not quite, Ratnoo said. “Ground control stations are needed when we have multiple drones in a specific area. UAS traffic management is an important area of research that is conducted globally.
