Autonomous Flight in Unknown Indoor Environments

This paper addresses the problem of autonomous navigation of a micro air vehicle (MAV) in GPS‐denied environments. We present experimental validation and analysis for our system that enables a quadrotor helicopter, equipped with a laser range finder sensor, to autonomously explore and map unstructured and unknown environments. The key challenge for enabling GPS‐denied flight of a MAV is that the system must be able to estimate its position and velocity by sensing unknown environmental structure with sufficient accuracy and low enough latency to stably control the vehicle. Our solution overcomes this challenge in the face of MAV payload limitations imposed on sensing, computational, and communication resources. We first analyze the requirements to achieve fully autonomous quadrotor helicopter flight in GPS‐denied areas, highlighting the differences between ground and air robots that make it difficult to use algorithms developed for ground robots. We report on experiments that validate our solutions to key challenges, namely a multilevel sensing and control hierarchy that incorporates a high‐speed laser scan‐matching algorithm, data fusion filter, high‐level simultaneous localization and mapping, and a goal‐directed exploration module. These experiments illustrate the quadrotor helicopter's ability to accurately and autonomously navigate in a number of large‐scale unknown environments, both indoors and in the urban canyon. The system was further validated in the field by our winning entry in the 2009 International Aerial Robotics Competition, which required the quadrotor to autonomously enter a hazardous unknown environment through a window, explore the indoor structure without GPS, and search for a visual target. © 2011 Wiley Periodicals, Inc.

Micro aerial vehicles (MAVs) have high potential for information gathering tasks to support situation awareness in search and rescue scenarios. Manually controlling MAVs in such scenarios requires experienced pilots and is error-prone, especially in stressful situations of real emergencies. The conditions of disaster scenarios are also challenging for autonomous MAV systems. The environment is usually not known in advance and GNSS might not always be available. We present a system for autonomous MAV flights in unknown environments which does not rely on global positioning systems. The method is evaluated in multiple search and rescue scenarios and allows for safe autonomous flights, even when transitioning between indoor and outdoor areas.

Enabling a flying robot to navigate autonomously in an unstructured and unknown indoor or outdoor environment is a big challenge. Attitude is one of the most important parameters for an autonomous UAV during such a flight. In this paper, we present the use of a vision system that can be used to estimate vehicle attitude using monocular pinhole camera only. Vision system are readily available on various UAV platforms and can be used for this purpose. Unlike previous work, this approach does not require (natural or artificial) parallel lines, ground markers or landmarks on the environment as clues, or use external sensors (e.g. VICON or GPS), additional on-board sensors (e.g. IMU), or additional special purpose cameras or multiple sensors. The proposed algorithm relies on on-board vision only, and the vision processing and control are performed on-board the vehicle using a single-board vision computer. After explanation of the algorithm, detailed presentation of the developed system and experimental set-up are provided. Finally, the results of the estimation and control experiments performed on a model quadrotor helicopter are presented.

In unknown environments, it is highly important to acquire efficiently the map of the unknown environments, which can be realized with swarm robot systems. However, swarm robots suffer from the lack of sensor data due to inevitably limited resources. Then, the individual maps of swarm robots degenerate as time goes, and thus the collective map of merging the individual maps may be inaccurate. This paper presents a feature-based map merging system with mesh networks to acquire individual maps efficiently and merge them accurately. The proposed system was tested with two small robots equipped with relatively low-cost sensors in unknown indoor environments. The experimental results showed that the proposed system acquired successfully individual maps and merged them accurately.

This paper describes the latest efforts to develop an Automated UAV Mission System (AUMS) for small vertical takeoff and landing (VTOL) unmanned air vehicles (UAVs). In certain applications such as force protection, perimeter security, and urban surveillance a VTOL UAV can provide far greater utility than fixed-wing UAVs or ground-based sensors. The VTOL UAV can operate much closer to an object of interest and can provide a hover-and-stare capability to keep its sensors trained on an object, while the fixed wing UAV would be forced into a higher altitude loitering pattern where its sensors would be subject to intermittent blockage by obstacles and terrain. The most significant disadvantage of a VTOL UAV when compared to a fixed-wing UAV is its reduced flight endurance. AUMS addresses this disadvantage by providing forward staging, refueling, and recovery capabilities for the VTOL UAV through a host unmanned ground vehicle (UGV), which serves as a launch/recovery platform and service station. The UGV has sufficient payload capacity to carry UAV fuel for multiple launch, recovery, and refuel iterations. The UGV also provides a highly mobile means of forward deploying a small UAV into hazardous areas unsafe for personnel, such as chemically or biologically contaminated areas. Teaming small UAVs with large UGVs can decrease risk to personnel and expand mission capabilities and effectiveness. There are numerous technical challenges being addressed by these development efforts. Among the challenges is the development and integration of a precision landing system compact and light enough to allow it to be mounted on a small VTOL UAV while providing repeatable landing accuracy to safely land on the AUMS. Another challenge is the design of a UGV-transportable, expandable, self-centering landing pad that contains hardware and safety devices for automatically refueling the UAV. A third challenge is making the design flexible enough to accommodate different types of VTOL UAVs, such as the AAI iSTAR ducted-fan vehicle and small helicopter UAVs. Finally, a common command-and-control architecture which supports the UAV, UGV, and AUMS must be developed and interfaced with these systems to allow fully autonomous collaborative behaviors. Funded by the Joint Robotics Program, AUMS is part of a joint effort with the Air Force Research Laboratory and the Army Missile Research Development and Engineering Command. The objective is to develop and demonstrate UGV-UAV teaming concepts and work with the warfighter to ensure that future upgrades are focused on operational requirements. This paper describes the latest achievements in AUMS development and some of the military program and first responder situations that could benefit from this system.

Indoor motion planning focuses on solving the problem of navigating an agent through a cluttered environment. To date, quite a lot of work has been done in this field, but these methods often fail to find the optimal balance between computationally inexpensive online path planning, and optimality of the path. Along with this, these works often prove optimality for single-start single-goal worlds. To address these challenges, we present a multiple waypoint path planner and controller stack for navigation in unknown indoor environments where waypoints include the goal along with the intermediary points that the robot must traverse before reaching the goal. Our approach makes use of a global planner (to find the next best waypoint at any instant), a local planner (to plan the path to a specific waypoint) and an adaptive Model Predictive Control strategy (for robust system control and faster maneuvers). We evaluate our algorithm on a set of randomly generated obstacle maps, intermediate waypoints and start-goal pairs, with results indicating significant reduction in computational costs, with high accuracies and robust control.

Robot odor source localization in indoor environments based on gradient adaptive extremum seeking search

This paper proposes a method for collaborative exploration adopting multiple UAVs in an unknown GPS-denied indoor environment. The core of this method is to use the Tracking-D*Lite algorithm to track moving targets in unknown terrain, combined with the Wall-Around algorithm based on the Bug algorithm to navigate the UAV in the unknown indoor environment. The method adopts the advantages of the above two algorithms, where the UAV applies the Wall-Around algorithm to fly around the wall and utilizes the Tracking-D*Lite algorithm to achieve collaboration among UAVs. This method is simulated and visualized by using Gazebo, and the results show that it can effectively take the advantages of multiple UAVs to explore the unknown indoor environments. Moreover, the method can also draw the boundary-contour map of the entire environment at last. Once extended to the real world, this method can be applied to dangerous buildings after earthquakes, hazardous gas factories, underground mines, or other search and rescue scenarios.

This paper presents our solution for enabling a quadrotor helicopter to autonomously navigate unstructured and unknown indoor environments. We compare two sensor suites, specifically a laser rangefinder and a stereo camera. Laser and camera sensors are both well-suited for recovering the helicopter's relative motion and velocity. Because they use different cues from the environment, each sensor has its own set of advantages and limitations that are complimentary to the other sensor. Our eventual goal is to integrate both sensors on-board a single helicopter platform, leading to the development of an autonomous helicopter system that is robust to generic indoor environmental conditions. In this paper, we present results in this direction, describing the key components for autonomous navigation using either of the two sensors separately.

The problem of cooperative obstacle avoidance by a group of unmanned ground vehicles (UGVs) and unmanned air vehicles (UAVs) during a typical search mission is addressed in this paper. The group of UAVs and UGVs are performing a search operation in the designated area. All the UAVs and UGVs are equipped with a vision sensor/LIDAR to identify the possible obstacles in the search space. Due to their operation on the ground, UGVs are more likely to encounter obstacles. The obstacle avoidance for UGV under the event of sensor failure is done with environment information from the nearest UAV. The UAV plans its trajectory according to the UGV’s expected future trajectory, leading towards the base station. The UGV replans its trajectory to avoid obstacles after obtaining the information from its nearest UAV. A simulation study is performed with 10 UAVs and five UGVs performing a search mission in 1[Formula: see text]km[Formula: see text][Formula: see text][Formula: see text]1[Formula: see text]km area. The proposed obstacle avoidance method is experimentally validated in the outdoor environment with an autonomous UAV equipped with a camera and an autonomous UGV navigating based on GPS localization and environment information from the UAV.

In this paper autonomous formation control for Unmanned Aerial Vehicles (UAVs) has been discussed and a real time solution has been put forward by benefiting General Purpose Graphical Processing Units (GPGPU) accelerated potential field approach while ensuring obstacle and collision avoidance in unknown environment by using real-time sensors. GPGPU accelerated real time formation control for UAVs was designed and the basic model of the approach has been explained in our previous work (Cetin and Yilmaz 2014). As the deficiencies of the previous approach, autonomous real time collision and coordinated obstacle avoidance features in unknown environments are also handled while maintaining formation flight conditions in this work. With these features, improved autonomous formation control approach is discussed as a real time solution. The computation is performed by using Graphical Processing Units (GPUs) as parallel computation architectures by benefiting from Single Instruction Multiple Data (SIMD) type parallel algorithms. Classic binary map conversation, connected component labeling and minimum bounding box algorithms which are commonly used for image processing applications, has been evaluated for real time obstacle detection and avoidance features by developing GPGPU suitable parallel algorithms. Real-time solution has been developed by integrated these parallel algorithms with parallel Artificial Potential Field (APF) computation algorithm. Simulation results are proved that this novel autonomous improved formation control approach is successful and it would be used in real time applications like UAV formation flight missions.

For Unmanned Air Vehicles (UAV) to operate autonomously in unstructured and GPS-denied environments, requisite information about the surroundings with adequate detailing needs to be generated and Simultaneous Localization and Mapping (SLAM) is among the most preferred protocols for fulfilling these needs. This study presents a SLAM-based linear optimal control approach for UAV navigation with limited sensor resources. The suggested protocol utilizes 1-Point RANSAC and an Extended Kalman Filter (EKF) for SLAM from a 6 degree-of-freedom motion monocular image sequence. Output of this research effort includes the estimated camera motion and a sparse map of salient point features through sensory representation. In this study we present a unique combination of 1-Point RANSAC (Random Sample Consensus) in conjunction with EKF with innate focus on reducing of the computational complexity. Contemporary studies by UAV research groups have successfully demonstrated the usefulness of algorithms based on 1-Point RANSAC and also provided comparison of algorithmic results with those of visual odometry. The present study further extends the scope by evaluating the algorithm with data generated from an input device mounted on a custom developed UAV. Employment of SLAM mission for mapping the areas that are prone to mining and land exploration is among the important outcomes of the proposed study.

This work presents a proposal for an Ambient Assisted Wheelchair (A2W). Such a vehicle could collect useful information provided by Intelligent Environments equipped with appropriate wireless services. Two different networks have been considered for this tasks: a standard Wi-Fi network that will allow a connection to Internet Services providing building accessibility information; and a specific sensor network able to provide indoor location. The Buildings Accessibility Information Service (BAIS) is aimed for providing route planning in unknown indoor environments. BAIS system is based on a set of XML description files that contain all the information needed to access public or private buildings: floor maps, accessibility information, available routes and calibration landmarks. Sensor Network (SN) is based on a set of wireless nodes equipped with ZigBee devices; RSSI (Received Signal Strength Indication) from fixed nodes are used to compute an approximate 3D location that reinforces vehicle positioning inside buildings. Both resources BAIS and SN (when installed) could provide valuable information to any kind of user inside unknown environments, although best performances will be achieved with A2W units.

Mobile robots require the ability to build their own maps to operate in unknown environments. This article introduces a new online method for learning world maps in non-static worlds. The new approach expands the previously proposed method by Araujo et al. (1998, 1999), based on the fuzzy ART neural architecture (FARTNA) towards the application in non-static worlds. For this purpose we introduce the prune-able fuzzy ART neural architecture (PAFARTNA) which extends the FARTNA self-organizing neural network to include the ability to selectively remove recognition categories. A method is proposed for the perception of object removals, and integrated with the PAFARTNA. To study and experimentally validate the proposed methods we integrated sensor information from a laser range finder sensor and a ring of infrared range sensors. Results of experiments with a Nomad 200 mobile robot are presented, demonstrating the effectiveness of the proposed methods.

Simultaneous localization and mapping (SLAM) has been richly researched in past years particularly with regard to range-based or visual-based sensors. Instead of deploying dedicated devices that use visual features, it is more pragmatic to exploit the radio features to achieve this task, due to their ubiquitous nature and the wide deployment of Wifi wireless network. In this paper, we present a novel approach for crowd-sensing simultaneous localization and radio fingerprint mapping (C-SLAM-RF) in large unknown indoor environments. The proposed system makes use of the received signal strength (RSS) from surrounding Wifi access points (AP) and the motion tracking data from a smart phone (Tango as an example). These measurements are captured duration the walking of multiple users in unknown environments without map information and location of the AP. The experiments were done in a university building with dynamic environment and the results show that the proposed system is capable of estimating the tracks of a group of users with an accuracy of 1.74 meters when compared to the ground truth acquired from a point cloud-based SLAM.