Fatemeh Rastgar

Papers

Total Citations

H-Index

About

Fatemeh Rastgar is pioneering advances in autonomous mobile robot navigation, with a focus on overcoming the fundamental challenge of non-convex trajectory optimization in unknown and dynamic environments. Her work addresses a critical bottleneck: existing local planners often become trapped in poor local optima, leading to collisions. Rastgar’s key contribution, the PRIEST framework (7 citations, 2024), introduces projection-guided sampling-based optimization that enables efficient navigation without requiring a feasible global path, dramatically improving robustness in real-world settings. She has further accelerated these solutions through GPU-accelerated batch trajectory optimization (3 citations, 2023), demonstrating that running multiple optimizer initializations in parallel can escape local minima and ensure safer navigation. Most notably, her work on fast joint multi-robot trajectory optimization (3 citations, 2022) achieves computation of trajectories for tens of aerial swarm robots in under a fraction of a second by breaking the joint optimization into decoupled sub-problems solved in parallel on GPUs. This breakthrough makes real-time coordination of large robot swarms computationally feasible for the first time. Rastgar’s research sits at the intersection of optimization theory, parallel computing, and practical robotics, offering scalable solutions that push the boundaries of autonomous navigation in complex environments.