Dhruva Kumar
Papers
2
Total Citations
4
H-Index
2
About
Dhruva Kumar is a roboticist working at the intersection of autonomous navigation, simultaneous localization and mapping (SLAM), and visual-language reasoning. His research addresses a critical bottleneck in field robotics: how to enable reliable, long-term autonomy on low-compute platforms with limited sensor fields of view. In his highly cited work on "Lighthouses and Global Graph Stabilization," Kumar introduced a novel active SLAM framework that uses sparse, re-observable landmarks—analogous to lighthouses—to stabilize global maps during exploration, preventing the catastrophic drift that plagues traditional SLAM on resource-constrained robots. This contribution is foundational for deploying robust autonomy in real-world environments where computational power is at a premium. More recently, with "VLPG-Nav," Kumar has advanced the frontier of object-goal navigation by integrating visual language models with pose graphs and probabilistic object localization maps. His method uniquely addresses the practical challenge of not just reaching a target object, but centering it within the robot’s camera frame—a critical capability for subsequent manipulation or inspection tasks. With over 4 citations across his most prominent works and growing recognition for bridging geometric SLAM with semantic understanding, Kumar is shaping the future of intelligent, perceptive robots that can operate reliably in the wild.
Research Focus
Key Achievements
Top Papers
- 1
- 2