There has been a considerable amount of recent work on high-speed micro-aerial vehicle flight in unknown and unstructured environments. Generally these approaches either use active sensing or fly slowly enough to ensure a safe braking distance with the relatively short sensing range of passive sensors. The former generally requires carrying large and heavy LIDARs and the latter only allows flight far away from the dynamic limits of the vehicle. One of the significant challenges for high-speed flight is the computational demand of trajectory planning at sufficiently high rates and length scales required in outdoor environments. We tackle both problems in this work by leveraging semantic information derived from an RGB camera on-board the vehicle. We first describe how to use semantic information to increase the effective range of perception on certain environment classes. Second, we present a sparse representation of the environment that is sufficiently lightweight for long distance path planning. We show how our approach outperforms more traditional metric planners which seek the shortest path, demonstrate the semantic planner's capabilities in a set of simulated and excessive real-world autonomous quadrotor flights in an urban environment.

There has been considerable recent work in motion planning for UAVs to enable aggressive, highly dynamic flight in known environments with motion capture systems. However, these existing planners have not been shown to enable the same kind of flight in unknown, outdoor environments. In this paper we present a receding horizon planning architecture that enables the fast replanning necessary for reactive obstacle avoidance by combining three techniques. First, we show how previous work in computationally efficient, closed-form trajectory generation method can be coupled with spatial partitioning data structures to reason about the geometry of the environment in real-time. Second, we show how to maintain safety margins during fast flight in unknown environments by planning velocities according to obstacle density. Third, our receding-horizon, sampling-based motion planner uses minimum-jerk trajectories and closed-loop tracking to enable smooth, robust, high-speed flight with the low angular rates necessary for accurate visual-inertial navigation. We compare against two state-of-the-art, reactive motion planners in simulation and benchmark solution quality against an offline global planner. Finally, we demonstrate our planner over 80 flights with a combined distance of 22km of autonomous quadrotor flights in an urban environment at speeds up to 9.4ms $^{-1}$.

Autonomous navigation through unknown environments is a challenging task that entails real-time localization, perception, planning, and control. UAVs with this capability have begun to emerge in the literature with advances in lightweight sensing and computing. Although the planning methodologies vary from platform to platform, many algorithms adopt a hierarchical planning architecture where a slow, low-fidelity global planner guides a fast, high-fidelity local planner. However, in unknown environments, this approach can lead to erratic or unstable behavior due to the interaction between the global planner, whose solution is changing constantly, and the local planner; a consequence of not capturing higher-order dynamics in the global plan. This work proposes a planning framework in which multi-fidelity models are used to reduce the discrepancy between the local and global planner. Our approach uses high-, medium-, and low-fidelity models to compose a path that captures higher-order dynamics while remaining computationally tractable. In addition, we address the interaction between a fast planner and a slower mapper by considering the sensor data not yet fused into the map during the collision check. This novel mapping and planning framework for agile flights is validated in simulation and hardware experiments, showing replanning times of 5-40 ms in cluttered environments.

Although unmanned air vehicles' increasing agility and autonomy may soon allow for flight in urban environments, the impact of complex urban wind fields on vehicle flight performance remains unclear. Unlike synoptic winds at high altitudes, urban wind fields are subject to turbulence generated by the buildings and terrain. The resulting spatial and temporal variation makes inference about the global wind field based on local wind measurements difficult and prevents the use of most simple wind models. Fortunately, the structure of the urban environment provides exploitable predictability given a suitable computational fluid dynamics solver, a representative 3D model of the environment, and an estimate of the expected prevailing wind speed and heading. The prevailing wind speed and direction at altitude and computational fluid dynamics solver can generate the corresponding wind field estimate over the map. By generating wind fields in this way, this work investigates a quadrotor's ability to exploit them for improved flight performance. Along with the wind field estimate, an empirically derived power consumption model is used to find minimum-energy trajectories with a planner both aware of and naive to the wind field. When compared to minimum-energy trajectories that do not incorporate wind conditions, the wind-aware trajectories demonstrate reduced flight times, total energy expenditures, and failures due to excess air speed for trajectories across MIT campus.

We present a motion planning algorithm for dynamic vehicles navigating through unknown environments. We focus on the scenario in which a fast-moving car attempts to navigate from a start location to a set of goal coordinates in minimum time with no prior information about the environment, building a map in real time from onboard sensor data. Whereas existing planners for exploration confine themselves to a conservative set of constraints to guarantee safety around unknown regions of the environment, we instead learn a hazard function from data, which maps the vehicle's dynamic state and current environment knowledge to a probability of collision. We perform receding horizon planning in which the objective function is evaluated in expectation over those learned probabilities of collision. Our algorithm demonstrates sensible emergent behaviors, like swinging wide around blind corners, slowing down near the map frontier, and accelerating in regions of high visibility. Our algorithm is capable of navigating from start to goal much more quickly than the conservative baseline planner without sacrificing safety. We demonstrate our algorithm on a 1:8-scale high-performance RC car equipped with a planar laser range-finder and inertial measurement unit, reaching speeds of 4m/s in unknown, indoor spaces. A video of experimental results is available at: http: //groups.csail.mit.edu/rrg/nav_learned_prob_collision.