Project Title: Emergent Collective Intelligence for Ephemeral Resource Tracking
(Agent-based modeling, reinforcement learning, and dynamic environments)
Supervisors:
• Dr. Akanksha Rathore – Assistant Professor, Computer Science, BITS Pilani, Hyderabad Campus
• Dr. Ariana Strandburg-Peshkin – Group Leader, Animal Behavior, Max Planck Institute of Animal
Behavior, Germany
Project Overview:
How do social animals adaptively locate transient, patchy resources like water in deserts, without centralized
coordination or maps? This interdisciplinary project explores how local interactions and learning rules give
rise to emergent collective intelligence in dynamic, resource-scarce environments.
Focusing on arid and semi-arid ecosystems, the project combines theories and tools from animal behavior,
reinforcement learning, agent-based modeling, and ecological data analysis. Simulations will integrate
empirically grounded social rules with adaptive learning, tested against real movement datasets (e.g.,
Movebank) and satellite-derived environmental dynamics.
This JRF position is part of a prestigious Indo-German Science & Technology Centre (IGSTC-WISER)
funded project with mentoring from both Indian and German institutions. The JRF will work in a team with
exposure to computational, ecological, and theoretical tools, with strong potential for PhD conversion. The
position also offers opportunities to co-mentor MSc students and to participate in collaborative visits to
Konstanz, Germany.
Project Objectives:
• Model social dynamics of resource tracking using agent-based simulations.
• Integrate reinforcement learning into agents for adaptive strategy development.
• Ground simulations in real environments using satellite data (e.g., NDVI, NDWI).
• Calibrate and validate agent rules using real animal tracking data.
• Compare social vs. RL-based strategies under uncertainty.
