2023 Author: Bryan Walter | [email protected]. Last modified: 2023-05-21 22:24
French engineers have created a robot that uses the methods of ants that live in deserts to navigate. The developers taught him to find his way home by counting steps, measuring optical flow while walking, and determining the direction by the radiation of the sun. Experiments have shown that the robot is able to successfully return to its original position by the shortest path, even if it previously moved randomly, according to an article published in Science Robotics.
Most modern devices and modes of transport use signals from satellite systems for accurate navigation. Since these signals may not be available, engineers are developing systems that can operate autonomously. For example, rockets, spacecraft and other maneuverable flying objects are often equipped with an inertial navigation system that calculates the distance traveled and direction based on the readings of gyroscopes and accelerometers. In addition, researchers continue to look, including in living organisms, for other ways of navigation.
One of the most famous animals with a reliable and versatile navigation system are ants. As a rule, they use the trail pheromones left behind while driving to find their way home. However, this method does not work in all conditions. For example, in hot deserts, pheromones released by ants are simply destroyed by sunlight and heat. Instead of marking the path with substances, desert ants such as Cataglyphis fortis and Melophorus bagoti use a combination of several methods. Over long distances, they use a path integration method to calculate the shortest path home. To do this, they count the steps taken, estimate the optical flux (the speed of movement of the surface relative to the retina) and calculate the direction along the height of the Sun and the polarization of the scattered radiation of the sky, which depends on the time of day. When an ant comes to the end of the path using path integration, it uses visual cues (the most prominent objects) to find the entrance to the nest.
Stephane Viollet and his colleagues at the University of Aix-Marseille have created a robot that uses techniques found in desert ants to navigate. It has six legs, as well as a variety of sensors that are collected in a Raspberry Pi single board computer. As a solar compass, the robot uses a pair of ultraviolet sensors, as well as polarizers, the rotation of which makes it possible to establish the distribution of the polarization of the incident light in the sky, and, therefore, to determine the direction of movement. In addition, an optical flow sensor is installed in the robot. It consists of two rows of six hexagonal pixels. Due to the delay in the appearance of the image on two adjacent pixels, the sensor can calculate the optical flow. Also, the robot is able to remember the number of steps taken.
Robot components
As an experimental site, the authors prepared a floor with a textured surface on which the robot walked. At first, he moved along a random curve, reaching the edge of the site, and then returned to its original position, stopping several times to recalculate the direction and distance, taking into account the new data. During the experiment, engineers tested five different algorithms using one or more components of ant navigation.
Trajectories of the robot's movement under the control of various algorithms
Experiments have shown that when using the most complete algorithm, the robot successfully returned to its original position with an average error of 0.67 percent. The authors note that during the experiments, the robot walked about 14 meters, so the results of the experiments cannot be called completely repeating the results of the movement of real ants. Considering size, speed, and distance traveled, a robot would have to travel more than 30 kilometers under similar conditions to catch up with ants traveling several hundred meters.
In 2016, German scientists showed that navigation by the number of steps and optical flow in ants works independently of each other. They experimented with foraging ants carrying worker ants. If a pair of ants were separated, the worker could find his way back to the anthill. If a bandage was put on the eyes of this ant, he could no longer find his way back.
