Gireeja Ranade

Carnegie Mellon University, Microsoft Research (United Kingdom), Microsoft (United States)

Papers

Total Citations

H-Index

About

Gireeja Ranade is a leading researcher at the intersection of robotics, autonomous decision-making, and information theory, with a core focus on enabling robots to operate intelligently under severe resource constraints. Her work fundamentally addresses the **budgeted information gathering problem**, where a robot must maximize the information it collects from an unknown environment while operating within a fixed fuel or energy budget. This challenge is critical for autonomous exploration, inspection, and search-and-rescue missions. Ranade’s major contributions lie in developing **data-driven and imitation learning frameworks** for planning. Her most cited work, “Learning to gather information via imitation” (22 citations), pioneers a method where robots learn optimal exploration strategies by imitating expert demonstrations, bypassing the computational intractability of traditional optimization. She further advanced the field with “No-regret replanning under uncertainty” (14 citations), which provides robust, online path planning algorithms for robots operating in latent environments modeled by Gaussian Processes, ensuring performance guarantees even when replanning is necessary. Her 2018 paper on “Data-driven planning via imitation learning” (5 citations) solidifies this approach, demonstrating how to learn complex, task-specific objectives—from collision-free navigation to complete area mapping—directly from data. Ranade’s work is pivotal for creating autonomous systems that are both efficient and reliable in the real world.