Kyle Vedder
Papers
2
Total Citations
21
H-Index
2
About
Kyle Vedder is a researcher advancing the frontiers of autonomous navigation and multi-agent systems, with a focus on making real-time decision-making computationally tractable for resource-constrained robots. His work bridges theoretical algorithm design and practical embedded deployment. Vedder’s most cited paper, “X*: Anytime Multi-Agent Path Finding for Sparse Domains using Window-Based Iterative Repairs” (2020, 15 citations), introduces a novel anytime algorithm that efficiently solves multi-agent pathfinding in sparse environments by iteratively repairing suboptimal solutions within a sliding time window—a critical contribution for dynamic, real-world robotics. Complementing this, his 2022 work “Sparse PointPillars: Maintaining and Exploiting Input Sparsity to Improve Runtime on Embedded Systems” (6 citations) tackles the computational bottleneck of 3D perception on mobile platforms. By recognizing that Bird’s Eye View representations are inherently sparse, Vedder engineered a high-performance object detector that preserves and exploits this sparsity, achieving significant runtime improvements without sacrificing accuracy. This work directly addresses the hardware limitations of drones, rovers, and other embedded systems. Through these contributions, Vedder has demonstrated a clear ability to identify and exploit structural properties of problems—whether in pathfinding or perception—to create practical, efficient solutions for autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2