Are we there yet?
Uncovering universal principles of Homing in the lab and nature
August 19, 2024 | Bharti Dharapuram
Researchers studied the statistical principles of homing by programming a foraging robot (left) in the lab to mimic the noise in the flight trajectories of homing pigeons (right; main image from Sasaki et al. (2022) under a CC BY-NC-ND license, inset image by David McCorquodale modified under a CC by 4.0 license).
Homing is ubiquitous in nature, where animals traveling in search of food, mates, or other resources find their way back home using environmental cues. Tracking the motion of a foraging robot in the lab, researchers have discovered that homing time does not increase beyond a critical level of noise in the search process. It is thus suggested that course corrections in the homing path do not necessarily increase the efficiency of homing beyond this criticality. This statistical basis of homing behaviour was described by a collaborative team of researchers using experiments, simulations, theory and animal tracking data.
Many animals show homing behaviour, where they reorient themselves to make course corrections in their path as they travel home. For example, a dung beetle rolling a ball of dung towards its destination performs a ‘dance’ on top of it to reorient itself in response to its surroundings. While scientists have probed the biological mechanisms driving homing behaviour, the statistical principles underlying its dynamics are not well understood. Since it is not possible to systematically vary parameters of animal motion in their natural habitat, lab-based robots offer an alternative tunable system to investigate homing dynamics. Additionally, movement trajectories of animals hold clues about how noise in the search paths influences the homing process. These approaches were used by researchers from the Indian Institute of Technology (IIT) Bombay, The Institute of Mathematical Sciences, Chennai (IMSc) and IIT Mandi to study the physics of homing.
The research team programmed a robot to perform active dynamics, where it leaves a home location to find a target object and returns home by sensing light intensity. “The beautiful thing about the robot is that it can be programmed to create a wide spectrum of stochastic motions. We engineered it to perform course correction, which is what homing animals do using environmental cues,” says Arnab Pal from IMSc, an author of the study. An important aspect of their study was identifying orientation as the parameter to reset during course correction. “We wanted to understand the statistical properties of the time it takes to come back home with regard to the number of course corrections,” Pal says. They found that homing time increases with randomness in the robot’s orientation but remains constant beyond a critical value. This determines a critical frequency of course correction, beyond which homing efficiency does not improve. The researchers reproduced these findings using computer simulations of active Brownian particles, and explained them using a simple theoretical model from first principles.
Strikingly, the researchers were able to use these theoretical predictions to explain the real world flight trajectories of homing pigeons. “A single homing pigeon makes many mistakes but a flock with many individuals has less error,” Pal says. They found that flock size influences how the orientation of the flight path varies through time, as predicted by their model. Varying the noise in the robot’s wheels can mimic the behaviour of homing pigeons of different flock sizes. “There is a critical level of noise beyond which homing time does not change, which tells us that animals can only make mistakes until a certain limit”, he adds.
Looking ahead, the research team wants to pursue many different ideas related to homing behaviour. “Along with our collaborators, we want to see how the robot responds to obstacles in its homing path,” Pal says. While ecologists and statistical physicists often work in isolation, the research group’s interest lies in the intersection and bridging the gap using tools ranging from physics to ecology. “The motivation is to study the navigation of foraging living entities and learning in homing paths in an attempt to reveal and explain universal statistical patterns”.
Reference: Paramanick, S., Biswas, A., Soni, H., Pal, A., & Kumar, N. (2024). Uncovering universal characteristics of homing paths using foraging robots. PRX Life, 2(3), 033007. https://doi.org/10.1103/PRXLife.2.033007
