Search

Today, even the most compute-and-power constrained robots can measure complex, high data-rate video and LIDAR sensory streams. Often, such robots, ranging from low-power drones to space and subterranean rovers, need to transmit high-bitrate sensory data to a remote compute server if they are uncertain or cannot scalably run complex perception or mapping tasks locally. However, today's representations for sensory data are mostly designed for human, not robotic, perception and thus often waste precious compute or wireless network resources to transmit unimportant parts of a scene that are unnecessary for a high-level robotic task. This paper presents an algorithm to learn task-relevant representations of sensory data that are co-designed with a pre-trained robotic perception model's ultimate objective. Our algorithm aggressively compresses robotic sensory data by up to 11 × more than competing methods. Further, it achieves high accuracy and robust generalization on diverse tasks including Mars terrain classification with low-power deep learning accelerators, neural motion planning, and environmental timeseries classification. [ABSTRACT FROM AUTHOR]

Safe deployment of autonomous robots in diverse scenarios requires agents that are capable of efficiently adapting to new environments while satisfying constraints. In this article, we propose a practical and theoretically justified approach to maintain safety in the presence of dynamics uncertainty. Our approach leverages Bayesian meta-learning with last-layer adaptation. The expressiveness of neural-network features trained offline, paired with efficient last-layer online adaptation, enables the derivation of tight confidence sets, which contract around the true dynamics as the model adapts online. We exploit these confidence sets to plan trajectories that guarantee the safety of the system. Our approach handles problems with high dynamics uncertainty, where reaching the goal safely is potentially initially infeasible, by first exploring to gather data and reduce uncertainty, before autonomously exploiting the acquired information to safely perform the task. Under reasonable assumptions, we prove that our framework guarantees the high-probability satisfaction of all constraints at all times jointly, i.e., over the total task duration. This theoretical analysis also motivates two regularizers of last-layer meta-learning models that improve online adaptation capabilities as well as performance by reducing the size of the confidence sets. We extensively demonstrate our approach in simulation and on hardware. [ABSTRACT FROM AUTHOR]

Abstract The objective of this paper is to present a full-stack, real-time motion planning framework for kinodynamic robots and then show how it is applied and demonstrated on a physical quadrotor system operating in a laboratory environment. The proposed framework utilizes an offline–online computation paradigm, neighborhood classification through machine learning, sampling-based motion planning with an optimal cost distance metric, and trajectory smoothing to achieve real-time planning for aerial vehicles. This framework accounts for dynamic obstacles with an event-based replanning structure and a locally reactive control layer that minimizes replanning events. The approach is demonstrated on a quadrotor navigating moving obstacles in an indoor space and stands as, arguably, one of the first demonstrations of full-online kinodynamic motion planning, with execution cycles of 3 Hz to 5 Hz. For the quadrotor, a simplified dynamics model is used during the planning phase to accelerate online computation. A trajectory smoothing phase, which leverages the differentially flat nature of quadrotor dynamics, is then implemented to guarantee a dynamically feasible trajectory. Highlights • Full-stack framework for path planning and obstacle avoidance for agile robots. • Demonstrated on a quadrotor capable of dodging a fencing blade while flying indoors. • Fusion of sampling-based planning, machine learning, and reactive control. [ABSTRACT FROM AUTHOR]