Shafeef Omar

Papers

Total Citations

H-Index

About

Shafeef Omar is a roboticist whose research focuses on enabling safer, more adaptive locomotion for legged robots, particularly quadrupeds. His work sits at the intersection of perception, control, and learning, aiming to bridge the gap between theoretical planning and real-world terrain traversal. In his highly cited paper "SafeSteps: Learning Safer Footstep Planning Policies for Legged Robots via Model-Based Priors" (2023, 3 citations), Omar introduced a novel framework that integrates a-priori safety information—such as kinematic feasibility and shin collision avoidance—directly into learned footstep planning policies. This approach allows robots to proactively avoid hazardous terrain, moving beyond reactive safety measures. His earlier work, "Fast Convex Visual Foothold Adaptation for Quadrupedal Locomotion" (2022, 2 citations), further advanced the field by developing a perception-based controller that rapidly approximates foothold adjustments, building on his prior Visual Foothold Adaptation (VFA) and Model Predictive Control (MPC) methods. By combining fast convex optimization with visual feedback, Omar’s contributions are paving the way for more robust and efficient autonomous navigation in complex environments, with direct implications for search-and-rescue and industrial inspection robotics.