Yoshiaki Kuwata

Safe Maritime Autonomous Navigation With COLREGS, Using Velocity Obstacles

This paper presents an autonomous motion planning algorithm for unmanned surface vehicles (USVs) to navigate safely in dynamic, cluttered environments. The algorithm not only addresses hazard avoidance (HA) for stationary and moving hazards, but also applies the International Regulations for Preventing Collisions at Sea (known as COLREGS, for COLlision REGulationS). The COLREGS rules specify, for example, which vessel is responsible for giving way to the other and to which side of the “stand-on” vessel to maneuver. Three primary COLREGS rules are considered in this paper: crossing, overtaking, and head-on situations. For autonomous USVs to be safely deployed in environments with other traffic boats, it is imperative that the USVs navigation algorithm obeys COLREGS. Furthermore, when other boats disregard their responsibility under COLREGS, the USV must fall back to its HA algorithms to prevent a collision. The proposed approach is based on velocity obstacles (VO) method, which generates a cone-shaped obstacle in the velocity space. Because VOs also specify on which side of the obstacle the vehicle will pass during the avoidance maneuver, COLREGS are encoded in the velocity space in a natural way. Results from several experiments involving up to four vessels are presented, in what we believe is the first on-water demonstration of autonomous COLREGS maneuvers without explicit intervehicle communication. We also show an application of this motion planner to a target trailing task, where a strategic planner commands USV waypoints based on high-level objectives, and the local motion planner ensures hazard avoidance and compliance with COLREGS during a traverse.

Stereo vision-based obstacle avoidance for micro air vehicles using disparity space

We address obstacle avoidance for outdoor flight of micro air vehicles. The highly textured nature of outdoor scenes enables camera-based perception, which will scale to very small size, weight, and power with very wide, two-axis field of regard. In this paper, we use forward-looking stereo cameras for obstacle detection and a downward-looking camera as an input to state estimation. For obstacle representation, we use image space with the stereo disparity map itself. We show that a C-space-like obstacle expansion can be done with this representation and that collision checking can be done by projecting candidate 3-D trajectories into image space and performing a z-buffer-like operation with the disparity map. This approach is very efficient in memory and computing time. We do motion planning and trajectory generation with an adaptation of a closed-loop RRT planner to quadrotor dynamics and full 3D search. We validate the performance of the system with Monte Carlo simulations in virtual worlds and flight tests of a real quadrotor through a grove of trees. The approach is designed to support scalability to high speed flight and has numerous possible generalizations to use other polar or hybrid polar/Cartesian representations and to fuse data from additional sensors, such as peripheral optical flow or radar.

360-Degree Visual Detection and Target Tracking on an Autonomous Surface Vehicle

This paper describes perception and planning systems of an autonomous sea surface vehicle (ASV) whose goal is to detect and track other vessels at medium to long ranges and execute responses to determine whether the vessel is adversarial. The Jet Propulsion Laboratory (JPL) has developed a tightly integrated system called CARACaS (Control Architecture for Robotic Agent Command and Sensing) that blends the sensing, planning, and behavior autonomy necessary for such missions. Two patrol scenarios are addressed here: one in which the ASV patrols a large harbor region and checks for vessels near a fixed asset on each pass and one in which the ASV circles a fixed asset and intercepts approaching vessels. This paper focuses on the ASVs central perception and situation awareness system, dubbed Surface Autonomous Visual Analysis and Tracking (SAVAnT), which receives images from an omnidirectional camera head, identifies objects of interest in these images, and probabilistically tracks the objects presence over time, even as they may exist outside of the vehicles sensor range. The integrated CARACaS/SAVAnT system has been implemented on U.S. Navy experimental ASVs and tested in on-water field demonstrations.

Safe maritime navigation with COLREGS using velocity obstacles

This paper presents a motion planning algorithm for Unmanned Surface Vehicles (USVs) to navigate safely in dynamic, cluttered environments. The proposed algorithm not only addresses Hazard Avoidance (HA) for stationary and moving hazards but also applies the International Regulations for Preventing Collisions at Sea (known as COLREGS). The COLREGS rules specify, for example, which vessel is responsible for giving way to the other and to which side of the “stand-on” vessel to maneuver. The three primary COLREGS rules were considered in this paper: crossing, overtaking, and head-on situations. For USVs to be safely deployed in environments with other traffic boats, it is imperative that the USVs navigation algorithm obey COLREGS. Note also that if other boats disregard their responsibility under COLREGS, the USV will still apply its HA algorithms to avoid a collision. The proposed approach is based on Velocity Obstacles, which generates a cone-shaped obstacle in the velocity space. Because Velocity Obstacles also specify which side of the obstacle the vehicle will pass during the avoidance maneuver, COLREGS are encoded in the velocity space very naturally. The algorithm is demonstrated via both simulation and on-water tests.

Enabling continuous planetary rover navigation through FPGA stereo and visual odometry

Safe navigation under resource constraints is a key concern for autonomous planetary rovers operating on extraterrestrial bodies. Computational power in such applications is typically constrained by the radiation hardness and energy consumption requirements. For example, even though the microprocessors used for the Mars Science Laboratory (MSL) mission rover are an order of magnitude more powerful than those used for the rovers on the Mars Exploration Rovers (MER) mission, the computational power is still significantly less than that of contemporary desktop microprocessors. It is therefore important to move safely and efficiently through the environment while consuming a minimum amount of computational resources, energy and time. Perception, pose estimation, and motion planning are generally three of the most computationally expensive processes in modern autonomy navigation architectures. An example of this is on the MER where each rover must stop, acquire and process imagery to evaluate its surroundings, estimate the relative change in pose, and generate the next mobility system maneuver [1]. This paper describes improvements in the energy efficiency and speed of planetary rover autonomous traverse accomplished by converting processes typically performed by the CPU onto a Field Programmable Gate Arrays (FPGA) coprocessor. Perception algorithms in general are well suited to FPGA implementations because much of processing is naturally parallelizable. In this paper we present novel implementations of stereo and visual odometry algorithms on a FPGA. The FPGA stereo implementation is an extension of [2] that uses random in linear out rectification and a higher-performance interface between the rectification, filter, and disparity stages of the stereo pipeline. The improved visual odometry component utilizes a FPGA implementation of a Harris feature detector and sum of absolute differences (SAD) operator. The FPGA implementation of the stereo and visual odometry functionality have demonstrated a performance improvement of approximately three orders of magnitude compared to the MER-class avionics. These more efficient perception and pose estimation modules have been merged with motion planning techniques that allow for continuous steering and driving to navigate cluttered obstacle fields without stopping to perceive. The resulting faster visual odometry rates also allow for wheel slip to be detected earlier and more reliably. Predictions of resulting improvements in planetary rover energy efficiency and average traverse speeds are reported. In addition, field results are presented that compare the performance of autonomous navigation on the Athena planetary rover prototype using continuous steering or driving and continuous steering and driving with GESTALT traversability analysis using the FPGA perception and pose estimation improvements.

Chance-constrained dynamic programming with application to risk-aware robotic space exploration

Existing approaches to constrained dynamic programming are limited to formulations where the constraints share the same additive structure of the objective function (that is, they can be represented as an expectation of the summation of one-stage costs). As such, these formulations cannot handle joint probabilistic (chance) constraints, whose structure is not additive. To bridge this gap, this paper presents a novel algorithmic approach for joint chance-constrained dynamic programming problems, where the probability of failure to satisfy given state constraints is explicitly bounded. Our approach is to (conservatively) reformulate a joint chance constraint as a constraint on the expectation of a summation of indicator random variables, which can be incorporated into the cost function by considering a dual formulation of the optimization problem. As a result, the primal variables can be optimized by standard dynamic programming, while the dual variable is optimized by a root-finding algorithm that converges exponentially. Error bounds on the primal and dual objective values are rigorously derived. We demonstrate algorithm effectiveness on three optimal control problems, namely a path planning problem, a Mars entry, descent and landing problem, and a Lunar landing problem. All Mars simulations are conducted using real terrain data of Mars, with four million discrete states at each time step. The numerical experiments are used to validate our theoretical and heuristic arguments that the proposed algorithm is both (i) computationally efficient, i.e., capable of handling real-world problems, and (ii) near-optimal, i.e., its degree of conservatism is very low.

Probabilistic motion planning of balloons in strong, uncertain wind fields

This paper introduces a new algorithm for probabilistic motion planning in arbitrary, uncertain vector fields, with emphasis on high-level planning for Montgolfieré balloons in the atmosphere of Titan. The goal of the algorithm is to determine what altitude—and what horizontal actuation, if any is available on the vehicle—to use to reach a goal location in the fastest expected time. The winds can vary greatly at different altitudes and are strong relative to any feasible horizontal actuation, so the incorporation of the winds is critical for guidance plans. This paper focuses on how to integrate the uncertainty of the wind field into the wind model and how to reach a goal location through the uncertain wind field, using a Markov decision process (MDP). The resulting probabilistic solutions enable more robust guidance plans and more thorough analysis of potential paths than existing methods.

A risk-constrained multi-stage decision making approach to the architectural analysis of planetary missions

This paper presents a novel risk-constrained multi-stage decision making approach to the architectural analysis of planetary rover missions. In particular, focusing on a 2018 Mars rover concept, which was considered as part of a potential Mars Sample Return campaign, we model the entry, descent, and landing (EDL) phase and the rover traverse phase as four sequential decision-making stages. The problem is to find a sequence of divert and driving maneuvers so that the rover drive is minimized and the probability of a mission failure (e.g., due to a failed landing) is below a user-specified bound. By solving this problem for several different values of the model parameters (e.g., divert authority), this approach enables rigorous, accurate and systematic trade-offs for the EDL system vs. the mobility system, and, more in general, cross-domain trade-offs for the different phases of a space mission. The overall optimization problem can be seen as a chance-constrained dynamic programming problem, with the additional complexity that 1) in some stages the disturbances do not have any probabilistic characterization, and 2) the state space is extremely large (i.e, hundreds of millions of states for trade-offs with high-resolution Martian maps). To this purpose, we solve the problem by performing an unconventional combination of average and minimax cost analysis and by leveraging high efficient computation tools from the image processing community. Preliminary trade-off results are presented.

Joint chance-constrained dynamic programming

This paper presents a novel joint chance-constrained dynamic programming algorithm, which explicitly bounds the probability of failure to satisfy given state constraints. Existing constrained dynamic programming approaches cannot handle a joint chance constraint since their application is limited to constraints in the same form as the cost function, that is, an expectation over a sum of one-stage costs. We overcome this challenge by reformulating the joint chance constraint into a constraint on an expectation over a sum of indicator functions, which can be incorporated into the cost function by dualizing the optimization problem. As a result, the primal variables can be optimized by a standard dynamic programming, while the dual variable is optimized by a root-finding algorithm that converges exponentially. Error bounds on the primal and dual objective values are rigorously derived. We demonstrate the algorithm on a path planning problem, as well as an optimal control problem for Mars entry, descent and landing. The simulations are conducted using a real terrain data of Mars, with four million discrete states at each time step.

Decomposition algorithm for global reachability analysis on a time-varying graph with an application to planetary exploration

Hot air (Montgolfiere) balloons represent a promising vehicle system for possible future exploration of planets and moons with thick atmospheres such as Venus and Titan. To go to a desired location, this vehicle can primarily use the horizontal wind that varies with altitude, with a small help of its own actuation. A main challenge is how to plan such trajectory in a highly nonlinear and time-varying wind field. This paper poses this trajectory planning as a graph search on the space-time grid and addresses its computational aspects. When capturing various time scales involved in the wind field over the duration of long exploration mission, the size of the graph becomes excessively large. We show that the adjacency matrix of the graph is block-triangular, and by exploiting this structure, we decompose the large planning problem into several smaller subproblems, whose memory requirement stays almost constant as the problem size grows. The approach is demonstrated on a global reachability analysis of a possible Titan mission scenario.