Abstract
Active-sensing animals like bats probe their environment by emitting energy and analyzing returning signals. A single bat's call generates a cascade of echoes from objects, which the auditory system must segregate into one coherent spatial scene. This problem is exaggerated in social groups as one must also discriminate own echoes from the calls and echoes of con-specifics. Then, how do bats navigate dense swarms without colliding when facing jamming? To answer this, we develop biologically parameterized models of bats. First, we use a static model to characterize the auditory scene of individual bats. We find that echo-acoustic information in collectives is limited and is constrained by the group’s orientation and individual's spatial position. We then construct a dynamic model to investigate collective coordination despite compromised sensing and heuristic behaviors that bats could be utilizing. We demonstrate a reduction in collision rates when individuals integrate information over larger times and respond to calls of others rather than relying on self-echoes. We also explore echo-acoustic flow, where bats may utilize the consistency in sound arrival and angle to enhance echo detection and mitigate jamming costs in dense groups.
About the speaker
I am a masters student at the Indian Institute for Science Education and Research, Pune (IISER Pune). My thesis work, supervised by Thejasvi Beleyur, focuses on developing agent based models to understand bat collective behavior. I love animals (and their behavior ofc!) and I am having fun using my computational modeling skills to understand how bats do cool stuff!