Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper addresses the challenge of safe obstacle avoidance in fast quadrotor flight by introducing a nonlinear, actuator-aware safety filter that operates within the full dynamics of the quadrotor, rather than relying on simplified models. By leveraging an analytic collision cone barrier for 3D Gaussian Splatting (3DGS), the authors derive a high-relative-degree collision cone exponential control barrier function (CBF) that significantly improves trajectory smoothness and computational efficiency. The proposed method demonstrates a 47% reduction in trajectory jerk and operates 2.25 times faster than existing state-of-the-art safety filters, validated through both simulation and real-time hardware tests in complex environments.
Achieving a 47% reduction in trajectory jerk while running 2.25 times faster than existing safety filters could redefine fast quadrotor navigation in cluttered environments.
Fast quadrotor flight requires safe obstacle avoidance under tight onboard compute limits. While 3D Gaussian Splatting (3DGS) provides a continuous, geometry-aware scene representation for perception-driven navigation, existing 3DGS safety filters use reduced-order models such as single- and double-integrators that ignore actuator limits and assume commanded accelerations are realized instantaneously. Building on an analytic collision cone barrier for 3DGS, we introduce a nonlinear, actuator-aware safety filter enforced through the full quadrotor dynamics. We derive a high-relative-degree collision cone exponential CBF and a backup CBF that preserves QP feasibility under input constraints using a forward-simulated backup policy. Compared with a state-of-the-art 3DGS safety filter, our approach reduces trajectory jerk by 47% and runs 2.25 times faster. We validate the method in simulation and on hardware for real-time navigation in cluttered, perception-derived environments.