Information-Driven Gas Distribution Mapping for Autonomous Mobile Robots

The ability to sense airborne pollutants with mobile robots provides a valuable asset for domains such as industrial safety and environmental monitoring. Oftentimes, this involves detecting how certain gases are spread out in the environment, commonly referred to as a gas distribution map, to subsequently take actions that depend on the collected information. Since the majority of gas transducers require physical contact with the analyte to sense it, the generation of such a map usually involves slow and laborious data collection from all key locations. In this regard, this paper proposes an efficient exploration algorithm for 2D gas distribution mapping with an autonomous mobile robot. Our proposal combines a Gaussian Markov random field estimator based on gas and wind flow measurements, devised for very sparse sample sizes and indoor environments, with a partially observable Markov decision process to close the robot's control loop. The advantage of this approach is that the gas map is not only continuously updated, but can also be leveraged to choose the next location based on how much information it provides. The exploration consequently adapts to how the gas is distributed during run time, leading to an efficient sampling path and, in turn, a complete gas map with a relatively low number of measurements. Furthermore, it also accounts for wind currents in the environment, which improves the reliability of the final gas map even in the presence of obstacles or when the gas distribution diverges from an ideal gas plume. Finally, we report various simulation experiments to evaluate our proposal against a computer-generated fluid dynamics ground truth, as well as physical experiments in a wind tunnel.

Autonomous robotic exploration with simultaneous environment and traversability models learning

In this study, we address generalized autonomous mobile robot exploration of unknown environments where a robotic agent learns a traversability model and builds a spatial model of the environment. The agent can benefit from the model learned online in distinguishing what terrains are easy to traverse and which should be avoided. The proposed solution enables the learning of multiple traversability models, each associated with a particular locomotion gait, a walking pattern of a multi-legged walking robot. We propose to address the simultaneous learning of the environment and traversability models by a decoupled approach. Thus, navigation waypoints are generated using the current spatial and traversability models to gain the information necessary to improve the particular model during the robot's motion in the environment. From the set of possible waypoints, the decision on where to navigate next is made based on the solution of the generalized traveling salesman problem that allows taking into account a planning horizon longer than a single myopic decision. The proposed approach has been verified in simulated scenarios and experimental deployments with a real hexapod walking robot with two locomotion gaits, suitable for different terrains. Based on the achieved results, the proposed method exploits the online learned traversability models and further supports the selection of the most appropriate locomotion gait for the particular terrain types.

Trajectory Planning and Optimization for Minimizing Uncertainty in Persistent Monitoring Applications

This paper considers persistent monitoring of environmental phenomena using unmanned aerial vehicles (UAVs). The objective is to generate periodic dynamically feasible UAV trajectories that minimize the estimation uncertainty at a set of points of interest in the environment. We develop an optimization algorithm that iterates between determining the observation periods for a set of ordered points of interest and optimizing a continuous UAV trajectory to meet the required observation periods and UAV dynamics constraints. The interest-point visitation order is determined using a Traveling Salesman Problem (TSP), followed by a greedy optimization algorithm to determine the number of observations that minimizes the maximum steady-state eigenvalue of a Kalman filter estimator. Given the interest-point observation periods and visitation order, a minimum-jerk trajectory is generated from a bi-level optimization, formulated as a convex quadratically constrained quadratic program. The resulting B-spline trajectory is guaranteed to be feasible, meeting the observation duration, maximum velocity and acceleration, region enter and exit constraints. The feasible trajectories outperform existing methods by achieving comparable observability at up to 47% higher travel speeds, resulting in lower maximum estimation uncertainty.

Optimal Maneuvering for Autonomous Vehicle Self-Localization

We consider the problem of optimal maneuvering, where an autonomous vehicle, an unmanned aerial vehicle (UAV) for example, must maneuver to maximize or minimize an objective function. We consider a vehicle navigating in a Global Navigation Satellite System (GNSS)-denied environment that self-localizes in two dimensions using angle-of-arrival (AOA) measurements from stationary beacons at known locations. The objective of the vehicle is to travel along the path that minimizes its position and heading estimation error. This article presents an informative path planning (IPP) algorithm that (i) uses the determinant of the self-localization estimation error covariance matrix of an unscented Kalman filter as the objective function; (ii) applies an l-step look-ahead (LSLA) algorithm to determine the optimal heading for a constant-speed vehicle. The novel algorithm takes into account the kinematic constraints of the vehicle and the AOA means of measurement. We evaluate the performance of the algorithm in five scenarios involving stationary and mobile beacons and we find the estimation error approaches the lower bound for the estimator. The simulations show the vehicle maneuvers to locations that allow for minimum estimation uncertainty, even when beacon placement is not conducive to accurate estimation.

Path Planning for Underwater Information Gathering Based on Genetic Algorithms and Data Stochastic Models

Recent technological developments have paved the way to the employment of Autonomous Underwater Vehicles (AUVs) for monitoring and exploration activities of marine environments. Traditionally, in information gathering scenarios for monitoring purposes, AUVs follow predefined paths that are not efficient in terms of information content and energy consumption. Informative Path Planning (IPP) represents a valid alternative, defining the path that maximises the gathered information. This work proposes a Genetic Path Planner (GPP), which consists in an IPP strategy based on a Genetic Algorithm, with the aim of generating a path that simultaneously maximises the information gathered and the coverage of the inspected area. The proposed approach has been tested offline for monitoring and inspection applications of Posidonia Oceanica (PO) in three different geographical areas. The a priori knowledge about the presence of PO, in probabilistic terms, has been modelled utilising a Gaussian Process (GP), trained on real marine data. The GP estimate has then been exploited to retrieve an information content of each position in the areas of interest. A comparison with other two IPP approaches has been carried out to assess the performance of the proposed algorithm.

SACBP: Belief space planning for continuous-time dynamical systems via stochastic sequential action control

We propose a novel belief space planning technique for continuous dynamics by viewing the belief system as a hybrid dynamical system with time-driven switching. Our approach is based on the perturbation theory of differential equations and extends sequential action control to stochastic dynamics. The resulting algorithm, which we name SACBP, does not require discretization of spaces or time and synthesizes control signals in near real-time. SACBP is an anytime algorithm that can handle general parametric Bayesian filters under certain assumptions. We demonstrate the effectiveness of our approach in an active sensing scenario and a model-based Bayesian reinforcement learning problem. In these challenging problems, we show that the algorithm significantly outperforms other existing solution techniques including approximate dynamic programming and local trajectory optimization.

Multi-robot informative path planning in unknown environments through continuous region partitioning

We consider the NP-hard problem of multirobot informative path planning in the presence of communication constraints, where the objective is to collect higher amounts of information of an ambient phenomenon. We propose a novel approach that uses continuous region partitioning into Voronoi components to efficiently divide an initially unknown environment among the robots based on newly discovered obstacles enabling improved load balancing between robots. Simulation results show that our proposed approach is successful in reducing the initial imbalance of the robots’ allocated free regions while ensuring close-to-reality spatial modeling within a reasonable amount of time.

Adaptive fovea for scanning depth sensors

Depth sensors have been used extensively for perception in robotics. Typically these sensors have a fixed angular resolution and field of view (FOV). This is in contrast to human perception, which involves foveating: scanning with the eyes’ highest angular resolution over regions of interest (ROIs). We build a scanning depth sensor that can control its angular resolution over the FOV. This opens up new directions for robotics research, because many algorithms in localization, mapping, exploration, and manipulation make implicit assumptions about the fixed resolution of a depth sensor, impacting latency, energy efficiency, and accuracy. Our algorithms increase resolution in ROIs either through deconvolutions or intelligent sample distribution across the FOV. The areas of high resolution in the sensor FOV act as artificial fovea and we adaptively vary the fovea locations to maximize a well-known information theoretic measure. We demonstrate novel applications such as adaptive time-of-flight (TOF) sensing, LiDAR zoom, gradient-based LiDAR sensing, and energy-efficient LiDAR scanning. As a proof of concept, we mount the sensor on a ground robot platform, showing how to reduce robot motion to obtain a desired scanning resolution. We also present a ROS wrapper for active simulation for our novel sensor in Gazebo. Finally, we provide extensive empirical analysis of all our algorithms, demonstrating trade-offs between time, resolution and stand-off distance.

Rapidly-Exploring Adaptive Sampling Tree*: A Sample-Based Path-Planning Algorithm for Unmanned Marine Vehicles Information Gathering in Variable Ocean Environments

This research presents a novel sample-based path planning algorithm for adaptive sampling. The goal is to find a near-optimal path for unmanned marine vehicles (UMVs) that maximizes information gathering over a scientific interest area, while satisfying constraints on collision avoidance and pre-specified mission time. The proposed rapidly-exploring adaptive sampling tree star (RAST*) algorithm combines inspirations from rapidly-exploring random tree star (RRT*) with a tournament selection method and informative heuristics to achieve efficient searching of informative data in continuous space. Results of numerical experiments and proof-of-concept field experiments demonstrate the effectiveness and superiority of the proposed RAST* over rapidly-exploring random sampling tree star (RRST*), rapidly-exploring adaptive sampling tree (RAST), and particle swarm optimization (PSO).

An informative path planning framework for UAV-based terrain monitoring

Unmanned aerial vehicles represent a new frontier in a wide range of monitoring and research applications. To fully leverage their potential, a key challenge is planning missions for efficient data acquisition in complex environments. To address this issue, this article introduces a general informative path planning framework for monitoring scenarios using an aerial robot, focusing on problems in which the value of sensor information is unevenly distributed in a target area and unknown a priori. The approach is capable of learning and focusing on regions of interest via adaptation to map either discrete or continuous variables on the terrain using variable-resolution data received from probabilistic sensors. During a mission, the terrain maps built online are used to plan information-rich trajectories in continuous 3-D space by optimizing initial solutions obtained by a coarse grid search. Extensive simulations show that our approach is more efficient than existing methods. We also demonstrate its real-time application on a photorealistic mapping scenario using a publicly available dataset and a proof of concept for an agricultural monitoring task.

Distributed Multi-Robot Information Gathering under Spatio-Temporal Inter-Robot Constraints

Information gathering (IG) algorithms aim to intelligently select the mobile robotic sensor actions required to efficiently obtain an accurate reconstruction of a physical process, such as an occupancy map, a wind field, or a magnetic field. Recently, multiple IG algorithms that benefit from multi-robot cooperation have been proposed in the literature. Most of these algorithms employ discretization of the state and action spaces, which makes them computationally intractable for robotic systems with complex dynamics. Moreover, they cannot deal with inter-robot restrictions such as collision avoidance or communication constraints. This paper presents a novel approach for multi-robot information gathering (MR-IG) that tackles the two aforementioned restrictions: (i) discretization of robot’s state space, and (ii) dealing with inter-robot constraints. Here we propose an algorithm that employs: (i) an underlying model of the physical process of interest, (ii) sampling-based planners to plan paths in a continuous domain, and (iii) a distributed decision-making algorithm to enable multi-robot coordination. In particular, we use the max-sum algorithm for distributed decision-making by defining an information-theoretic utility function. This function maximizes IG, while fulfilling inter-robot communication and collision avoidance constraints. We validate our proposed approach in simulations, and in a field experiment where three quadcopters explore a simulated wind field. Results demonstrate the effectiveness and scalability with respect to the number of robots of our approach.

UAV-IoT for Next Generation Virtual Reality

We investigate UAV-IoT data capture and networking for remote scene virtual reality (VR) immersion. We characterize the delivered immersion fidelity as a function of the assigned UAV-IoT capture/network rates and study the optimization problem of maximizing it, for given system/application constraints. We explore fast reinforcement learning to discover the best dynamic UAV-IoT network placement over the scene of interest to maximize the expected remote immersion fidelity. We design scalable source-channel viewpoint coding to maximize the expected reconstruction fidelity of the data captured at every UAV location at the ground-based aggregation point. Finally, we explore layered directional networking and rate-distortion-power optimized embedded scheduling methods to effectively transmit the encoded data and overcome network transients that lead to packet buffering, which represent the fourth system component of our framework. Experimental results demonstrate considerable performance efficiency gains enabled by each system component over the respective state-of-the-art reference methods, in delivered VR immersion fidelity, application interactivity/play-out latency, and transmission power consumption.

Learning Environmental Field Exploration with Computationally Constrained Underwater Robots: Gaussian Processes Meet Stochastic Optimal Control

Autonomous exploration of environmental fields is one of the most promising tasks to be performed by fleets of mobile underwater robots. The goal is to maximize the information gain during the exploration process by integrating an information-metric into the path-planning and control step. Therefore, the system maintains an internal belief representation of the environmental field which incorporates previously collected measurements from the real field. In contrast to surface robots, mobile underwater systems are forced to run all computations on-board due to the limited communication bandwidth in underwater domains. Thus, reducing the computational cost of field exploration algorithms constitutes a key challenge for in-field implementations on micro underwater robot teams. In this work, we present a computationally efficient exploration algorithm which utilizes field belief models based on Gaussian Processes, such as Gaussian Markov random fields or Kalman regression, to enable field estimation with constant computational cost over time. We extend the belief models by the use of weighted shape functions to directly incorporate spatially continuous field observations. The developed belief models function as information-theoretic value functions to enable path planning through stochastic optimal control with path integrals. We demonstrate the efficiency of our exploration algorithm in a series of simulations including the case of a stationary spatio-temporal field.

Sampling-based incremental information gathering with applications to robotic exploration and environmental monitoring

We propose a sampling-based motion-planning algorithm equipped with an information-theoretic convergence criterion for incremental informative motion planning. The proposed approach allows dense map representations and incorporates the full state uncertainty into the planning process. The problem is formulated as a constrained maximization problem. Our approach is built on rapidly exploring information-gathering algorithms and benefits from the advantages of sampling-based optimal motion-planning algorithms. We propose two information functions and their variants for fast and online computations. We prove an information-theoretic convergence for an entire exploration and information-gathering mission based on the least upper bound of the average map entropy. A natural automatic stopping criterion for information-driven motion control results from the convergence analysis. We demonstrate the performance of the proposed algorithms using three scenarios: comparison of the proposed information functions and sensor configuration selection, robotic exploration in unknown environments, and a wireless signal strength monitoring task in a lake from a publicly available dataset collected using an autonomous surface vehicle.

Robotic Active Information Gathering for Spatial Field Reconstruction with Rapidly-Exploring Random Trees and Online Learning of Gaussian Processes

Information gathering (IG) algorithms aim to intelligently select a mobile sensor actions required to efficiently obtain an accurate reconstruction of a physical process, such as an occupancy map, or a magnetic field. Many recent works have proposed algorithms for IG that employ Gaussian processes (GPs) as underlying model of the process. However, most algorithms discretize the state space, which makes them computationally intractable for robotic systems with complex dynamics. Moreover, they are not suited for online information gathering tasks as they assume prior knowledge about GP parameters. This paper presents a novel approach that tackles the two aforementioned issues. Specifically, our approach includes two intertwined steps: (i) a Rapidly-Exploring Random Tree (RRT) search that allows a robot to identify unvisited locations, and to learn the GP parameters, and (ii) an RRT*-based informative path planning that guides the robot towards those locations by maximizing the information gathered while minimizing path cost. The combination of the two steps allows an online realization of the algorithm, while eliminating the need for discretization. We demonstrate that our proposed algorithm outperforms state-of-the-art both in simulations, and in a lab experiment in which a ground-based robot explores the magnetic field intensity within an indoor environment populated with obstacles.

Spatial Spectroscopic Models for Remote Exploration

Ancient hydrothermal systems are a high-priority target for a future Mars sample return mission because they contain energy sources for microbes and can preserve organic materials (Farmer, 2000 ; MEPAG Next Decade Science Analysis Group, 2008 ; McLennan et al., 2012 ; Michalski et al., 2017 ). Characterizing these large, heterogeneous systems with a remote explorer is difficult due to communications bandwidth and latency; such a mission will require significant advances in spacecraft autonomy. Science autonomy uses intelligent sensor platforms that analyze data in real-time, setting measurement and downlink priorities to provide the best information toward investigation goals. Such automation must relate abstract science hypotheses to the measurable quantities available to the robot. This study captures these relationships by formalizing traditional "science traceability matrices" into probabilistic models. This permits experimental design techniques to optimize future measurements and maximize information value toward the investigation objectives, directing remote explorers that respond appropriately to new data. Such models are a rich new language for commanding informed robotic decision making in physically grounded terms. We apply these models to quantify the information content of different rover traverses providing profiling spectroscopy of Cuprite Hills, Nevada. We also develop two methods of representing spatial correlations using human-defined maps and remote sensing data. Model unit classifications are broadly consistent with prior maps of the site's alteration mineralogy, indicating that the model has successfully represented critical spatial and mineralogical relationships at Cuprite. Key Words: Autonomous science-Imaging spectroscopy-Alteration mineralogy-Field geology-Cuprite-AVIRIS-NG-Robotic exploration. Astrobiology 18, 934-954.

Active sensing for motion planning in uncertain environments via mutual information policies

This paper addresses path planning with real-time reaction to environmental uncertainty. The environment is represented as a robotic roadmap, or graph, and is uncertain in that the edges of the graph are unknown to the robot a priori. Instead, the robot’s prior information consists of a distribution over candidate edge sets, modeling the likelihood of certain obstacles in the environment. The robot can locally sense the environment, and at a vertex, can determine the presence or absence of some subset of edges. Within this model, the reactive planning problem provides the robot with a start location and a goal location and asks it to compute a policy that minimizes the expected travel and observation cost. In contrast to computing paths that maximize the probability of success, we focus on complete policies (i.e., policies that are guaranteed to navigate the robot to the goal or determine no such path exists). We prove that the problem is NP-hard and provide a suboptimal, but computationally efficient solution. This solution, based on mutual information, returns a complete policy and a bound on the gap between the policy’s expected cost and the optimal. We test the performance of the policy and the lower bound against that of the optimal policy and explore the effects of errors in the robot’s prior information on performance. Simulations are run on a flexible factory scenario to demonstrate the scalability of the proposed approach. Finally, we present a method to extend this solution to robots with faulty sensors.

Dec-MCTS: Decentralized planning for multi-robot active perception

We propose a decentralized variant of Monte Carlo tree search (MCTS) that is suitable for a variety of tasks in multi-robot active perception. Our algorithm allows each robot to optimize its own actions by maintaining a probability distribution over plans in the joint-action space. Robots periodically communicate a compressed form of their search trees, which are used to update the joint distribution using a distributed optimization approach inspired by variational methods. Our method admits any objective function defined over robot action sequences, assumes intermittent communication, is anytime, and is suitable for online replanning. Our algorithm features a new MCTS tree expansion policy that is designed for our planning scenario. We extend the theoretical analysis of standard MCTS to provide guarantees for convergence rates to the optimal payoff sequence. We evaluate the performance of our method for generalized team orienteering and online active object recognition using real data, and show that it compares favorably to centralized MCTS even with severely degraded communication. These examples demonstrate the suitability of our algorithm for real-world active perception with multiple robots.

Tongue-driven sonar beam steering by a lingual-echolocating fruit bat

Animals enhance sensory acquisition from a specific direction by movements of head, ears, or eyes. As active sensing animals, echolocating bats also aim their directional sonar beam to selectively “illuminate” a confined volume of space, facilitating efficient information processing by reducing echo interference and clutter. Such sonar beam control is generally achieved by head movements or shape changes of the sound-emitting mouth or nose. However, lingual-echolocating Egyptian fruit bats, Rousettus aegyptiacus, which produce sound by clicking their tongue, can dramatically change beam direction at very short temporal intervals without visible morphological changes. The mechanism supporting this capability has remained a mystery. Here, we measured signals from free-flying Egyptian fruit bats and discovered a systematic angular sweep of beam focus across increasing frequency. This unusual signal structure has not been observed in other animals and cannot be explained by the conventional and widely-used “piston model” that describes the emission pattern of other bat species. Through modeling, we show that the observed beam features can be captured by an array of tongue-driven sound sources located along the side of the mouth, and that the sonar beam direction can be steered parsimoniously by inducing changes to the pattern of phase differences through moving tongue location. The effects are broadly similar to those found in a phased array—an engineering design widely found in human-made sonar systems that enables beam direction changes without changes in the physical transducer assembly. Our study reveals an intriguing parallel between biology and human engineering in solving problems in fundamentally similar ways.

It is well known that animals move their eyes, ears, and heads towards stimuli of interest to selectively gather information in complex environments. Interestingly, lingual-echolocating fruit bats, which generate sonar signals for object localization by clicking their tongues, can rapidly switch the direction of the sonar beam without changing head aim or mouth shape. The mechanism underlying this capability has intrigued scientists and engineers alike. In this study, we used a combination of experimental measurements and theoretical modeling to solve this mystery. We discovered that the focus of this bat’s sound beam shifts systematically across a range of angles as the sonar frequency increases. This unusual multi-frequency structure can be captured by modeling the sound emission as an array of sound sources located along the side of the mouth and driven by the clicking tongue. Changing only the position of the tongue in this model can steer the sonar beam in different directions, showing an effect broadly similar to that found in a human-made sonar phased array—a design that enables changing beam direction without changing the physical transducer assembly. Our study thus reveals an intriguing parallel between biology and human engineering, which arrived at fundamentally similar solutions to the same problem.

Multipass Target Search in Natural Environments

Consider a disaster scenario where search and rescue workers must search difficult to access buildings during an earthquake or flood. Often, finding survivors a few hours sooner results in a dramatic increase in saved lives, suggesting the use of drones for expedient rescue operations. Entropy can be used to quantify the generation and resolution of uncertainty. When searching for targets, maximizing mutual information of future sensor observations will minimize expected target location uncertainty by minimizing the entropy of the future estimate. Motion planning for multi-target autonomous search requires planning over an area with an imperfect sensor and may require multiple passes, which is hindered by the submodularity property of mutual information. Further, mission duration constraints must be handled accordingly, requiring consideration of the vehicle’s dynamics to generate feasible trajectories and must plan trajectories spanning the entire mission duration, something which most information gathering algorithms are incapable of doing. If unanticipated changes occur in an uncertain environment, new plans must be generated quickly. In addition, planning multipass trajectories requires evaluating path dependent rewards, requiring planning in the space of all previously selected actions, compounding the problem. We present an anytime algorithm for autonomous multipass target search in natural environments. The algorithm is capable of generating long duration dynamically feasible multipass coverage plans that maximize mutual information using a variety of techniques such as ϵ-admissible heuristics to speed up the search. To the authors’ knowledge this is the first attempt at efficiently solving multipass target search problems of such long duration. The proposed algorithm is based on best first branch and bound and is benchmarked against state of the art algorithms adapted to the problem in natural Simplex environments, gathering the most information in the given search time.