Ninth Symposium on Indoor Flight Issues

The 2017 Symposium on Indoor Flight Issues was held during July 2017 at the American Venue in Atlanta, Georgia, USA, and in September 2017 at the Asia/Pacific Venue in Beijing, China. Links to the .pdf versions of the Symposium papers that met the formatting requirements are provided below:
Unmanned Aerial Vehicle of Team IFOR for the International Aerial Robotics Competition 2017
Saptadeep Debnath, Anudeepsekhar Bolimera, Srinath H. Rao, Ayush Shirsat, Parv Khandelwal, Sarthak Bhooshan, Sujith Sizon, Bhuvesh Gaind, Shaleen Bengani, Navya Zaveri, Sachin Samuel, V. Kalaichelvi
Birla Institute of Technology and Science, Pilani – Dubai Campus
Dubai, United Arab Emirates
This paper describes IFOR (Intelligent Flying Object for Reconnaissance), an autonomous aerial vehicle that has been developed by Team IFOR at BITS Pilani, Dubai Campus. The vehicle being discussed in this paper is able to localize and navigate in an indoor GPS denied sterile environment. Downward facing Optical Flow sensor Px4Flow is used to localize and stabilize the attitude (pitch, roll and yaw) and altitude of the vehicle using PID controllers. Ground robot detection is carried out by two inclined forward facing cameras and uses a LIDAR to avoid obstacle ground robots by setting the area near the paths of obstacle ground robots as hazardous area.
Low Cost Drone for Autonomous GPS Denied Navigation, Awareness, and Interaction
John Marcus
Embry-Riddle Aeronautical University
Daytona Beach, Florida, United States
The Robotics Association of Embry-Riddle has continued development on a low cost, GPS denied multirotor system. This system is capable of interacting with ground robots and avoiding dynamic obstacles. GPS denied systems are complicated due to the complexity of navigation without fixed reference points. Autonomy is achieved using two computers, a Pixracer with PX4 firmware, and an Odroid XU4, which communicate and mesh information from a combination of instruments, mainly an optical flow sensor, USB camera, and LIDAR. The optical flow camera is used in a feedback loop in order for the system to get velocity feedback, which is then meshed with other information on the onboard computers to achieve stable flight. The system then uses the USB camera to identify and interact with its objectives. The total cost for autonomy for this system is less than $600, this only includes the price of computers, sensors and wiring.
Technical Paper of HITCSC Team for The International Aerial Robotics Competition
Siqiang Wang, Qingtao Yu, Feng Li
Harbin Institute of Technology
Harbin, Heilongjiang, China
This manuscript introduces the scheme and progress of HITCSC Team for the International Aerial Robotics Competition in 2017. To present our scheme, we mainly depict the physical configuration and the guidance, control and navigation approach. To improve the ability of obstacle avoidance, a combination perception system of camera and lidar is used. Motion estimation technique and deep network based detection model are made use of to sense the status of the aerial robot and the environment around. A hierarchical control block is applied and random planning algorithm is used to solve the guidance and interaction problem. In final, the latest progress is illustrated.
Super Squadron technical paper for International Aerial Robotics Competition 2017
C. Aasish, S. Jayadeep, N. Gowri
Hindustan University
Chennai, Tamil Nadu, India
The Team Reconnaissance is a team from India presents an indigenously developed Super Squadron an autonomous aerial robot as the candidate for the 7th mission of the AUVSI International Aerial Robotics Competition (IARC).This paper describes the technology used in the aerial system for performing the required tasks of the mission. The robot is programmed to track and interact with the ground robot autonomously and this bot programmed to work like a like a sheepherder and thereby it controls and guides the ground robots. Visual navigation method is used as an alternative for the conventional GPS method of navigation. The robotic platforms are equipped with optical flow navigation sensors and LIDAR to navigate efficiently through the arena without hitting the obstacles. Guidance for ground robots is accomplished with effective multiple objects tracking algorithm. The robust design of this copter makes it a tough contender in this event.
Team Description Paper of Indian Institute of Technology Kharagpur for IARC
Somesh Kumar, Gaurav Gardi, Manash Pratim Das, Krishnakant Deshmukh, Ashwary Anand, Amit Pathak, Aman Chandra, Vivek Mudgal, Shivang Agrawal
Indian Institute of Technology Kharagpur
Kharagpur, West Bengal, India
This paper reports the current preparation strategy of Aerial Robotics Kharagpur, participating in IARC Mission 7 2017. Our main goal includes robust, indoor localization in GPS denied environments supported by optical flow sensors. Other features like ground i-robots detection and differentiation between the target bots and patrol bots is also discussed, followed by a brief description of the herding algorithm AI algorithm to be used.
Kennesaw State University Multirotor Aircraft for 2017 IARC
Albert Cheng, Steffen Lim, Alec Graves, Owen Ortmann, Luke Appling, Hector Saldana, Thomas Smith, Thomas Fagan, Joshua Hunter, Kyle Pawlowski, Robert Zenko
Kennesaw State University
Marietta, Georgia, United States
Kennesaw State Universitys Aerial Robotics Competition Team has developed a multirotor aerial vehicle to compete in the International Aerial Robotics Competition (IARC). The multirotor is designed to safely navigate through dynamic obstacles and interact with ground robots autonomously. The robust design allows the multirotor to safely withstand impacts such as ground robot interactions and small falling impacts. Additionally, the craft uses neural networking techniques to identify and track ground robots with more accuracy.
International Aerial Robotics Competition Technical Paper
Jonas Buxton, Mark Raymond Jr., Logan Thomure, Justin York
Missouri University of Science and Technology
Rolla, Missouri, United States
This report details our complete aerial robot’s hardware and software design, from our current progress to our goals for the near future.
Unmanned Hexrcopter of NUAA for 2017 International Aerial Robotics Competition
Sun Peng, Yang Yurong, Yin Yanqing, Li Jiahuan, Zeng Xu, Xu Yue
Nanjing University of Aeronautics and Astronautics
Nanjing, Jiangsu, China
This paper briefly introduces the strategy of NUAA team for the 2017 International Aerial Robotics Competition. We propose a complete system in detail which contains aerial robot, flight control system, vision navigation system, obstacle-avoidance system and communication part. The flight control system used for this competition is researched and developed totally by ourselves, which is important for the attitude control. For completing this task, the visual navigation is indispensable. A technique that could be used to solve the navigation problem in Mission 7 is “optical flow.” Vision technology is elaborated in this paper. Moreover, the precautions for the operations of hexrcopter are described in detail to prevent the occurrence of a sudden in the case of an emergency.
"Eagle Eye" Aerial Robot Team Technical Paper
Lifugang, Haozhiyang, Yaojiale
Ningxia University
Yinchuan, China
The International Aerial Robotics Competition was founded in 1991, the first time in 2012, the establishment of the Asia-Pacific Division in China, with the US Division simultaneously. The purpose of the contest is to promote the development of aerial robot technology, its tasks are the current unmanned aerial vehicle technology cannot be achieved, In accordance with the rules of the game, the air robots need to complete the very challenging task, The 7th generation task is on the UAV Indoor autonomous navigation presents new challenges. According to the task of the competition, the intelligent four-rotor platform is built. The platform uses the onboard processor to realize the upper function of positioning and navigation, and equipped with a variety of sensors, the optical flow sensor and IMU data fusion to achieve the positioning hover, Identification and navigation control through the monocular identification of the target, teh use of PID controller for tracking and driving a set of programs, experiments show the feasibility of the program.
Designing an Autonomous Aerial Robot for Successful Operations in a Non-Deterministic Indoor Environment
Abdullah Almosalami, Tyler Blanchard, Brady Goenner, Andrew Jones, Jeremy Straub
North Dakota State University
Fargo, North Dakota, United States
This paper provides an overview of the design of an aerial robot (also commonly known as an unmanned aerial vehicle or UAV) to meet the goals identified for Mission 7 of the International Aerial Robotics Competition. This mission requires the robot to engage in a herding application to influence ground robots to exit the playing field via a designated side. To meet this challenge, a hardware-software system was developed which implemented all elements required for flight, as well as software for target identification, decision making, and path planning. In this paper, particular focus is placed on the solutions related to the guidance, navigation, control, and autonomy requirements. An overview of the systems of the aerial robot is presented and the solutions developed for the mission challenges are discussed.
Technical Paper for the International Aerial Robotics Competition
Matias Christensen, Håkon Flatval, Magnus Reiersen, Martin Sollie, Christian Wilhelmsen
Norwegian University of Science and Technology
Trondheim, Norway
The aim of this paper is to describe a system for a fully autonomous MAV capable of solving the seventh IARC mission. The system is designed to perform on-line strategic planning, collision avoidance, robot-to-robot interaction and navigation, relying on INS sensors and camera vision in a GPS-denied environment.
Autonomous Navigation of a Multirotor using Visual Odometry and Dynamic Obstacle Avoidance
Pranav Singhania, Siddharth R N, Sovan Das, Akshay Kalkunte Suresh
PES University
Bangalore, Karnataka, India
In the present day, aerial vehicles such as micro fixed-wings or multirotor systems are important assets for industrial growth. These vehicles are generally autonomous or semi-autonomous and have a wide field of applications in military, mining or forest surveillance. All these applications are highly dependent on GPS (Global Positioning System) localization. Our project involves the design, assembly and configuration of a fully autonomous multirotor capable of manoeuvring in a GPS denied environment without the aid of large stationary points of reference. The multirotor is capable of interacting with external autonomous moving objects using a combination of control and computer vision algorithms. It creates efficient routes in flight for navigation whilst avoiding obstacles during the motion.
PCC's Autonomous Air Vehicle System for IARC 2017
Stephanie Guzman, Alberto Heras, Stephen Jewell, Colin LaSharr, Ryan Province, Frank Manning
Pima Community College
Tucson, Arizona, United States
The Pima Community College UAV Club has designed an air vehicle system to compete in the International Aerial Robotics Competition (IARC). The rules require an autonomous air vehicle to herd a group of 10 ground robots while avoiding collisions with a second group of 4 obstacle robots. All 14 ground-based robots are themselves autonomous and move according to their own internal algorithms, including specified responses to external collisions and mechanical forces. The air vehicle is designed to use machine vision as well as lidar and sonar scanning to sense the positions of ground robots, and to navigate relative to a 20 m x 20 m arena. The arena is marked with a known grid pattern
Autonomous Quadrotor for the 2017 IARC by Team Elikos
Christophe Bédard, Riad Mohamed Gahlouz, Pierre-Yves Lajoie, Arnaud Paré-Vogt, Justine Pepin, Eva Terriault
École Polytechnique de Montréal
Montréal, Quebec, Canada
The aim of this paper is to present team Elikos' system design for a fully autonomous vehicle capable of solving IARC mission 7a. Using various sensors including GNC sensors, cameras, and lasers, the vehicle shall be able of robot interaction while avoiding obstacles, and present autonomous navigation capabilities based on computer vision and sensors analytics work. This paper emphasises the work of team Elikos for the 2016-2017 year.
Autonomous Herding of Ground Robots Using Computer Vision State Estimation and Tracking with an Aerial Robot
Kevin Sheridan, Logesh Roshan Ramadoss, Jason King Ching Lo
Purdue University
West Lafayette, Indiana, United States
In this paper, we describe our system for Mission 7 of the International Aerial Robotics Competition. The goal of Mission 7 is to autonomously herd 10 ground robots by touch over one of the four edges of a 20 meter by 20 meter arena in a GPS denied environment while avoiding moving obstacles. Our solution uses a neural network to determine which ground robot to touch and how many times to touch it. We use a Mini-ITX motherboard with an Intel I5 and ROS as our main computer onboard the robot. Our robot has 5 cameras, 1 LiDAR, and 1 IMU. We localize by fusing visual odometry, a particle filter, and IMU readings in an EKF. We use a minimum time polynomial trajectory generator to feasible paths for execution with a closed loop controller.
Autonomous Robot Herding Through Physical Interaction
Jordan Liske, Jared McMaster, Anthony Morast, Aubrey Olson, Jacob Oursland, Samuel Vinella, Michael vander Wal
South Dakota School of Mines and Technology
Rapid City, South Dakota, United States
The South Dakota School of Mines and Technology Unmanned Aerial Vehicle Team’s entry in the 2017 Association for Unmanned Vehicle Systems International International Aerial Robotics Competition is a ducted fan quadrotor helicopter leveraging commercial off-the-shelf hardware and open source software. Localization and mapping is performed through a single machine vision camera using the Direct Sparse Odometry algorithm. Path planning is achieved through a custom goal algorithm and the MoveIt! planning framework. Risk management and fail-safe mechanisms have been implemented at all levels within the control system to prevent undesirable collisions and damage.
Autonomous Quadrotor for the 2017 International Aerial Robotics Competition
Rui Yang, Ziliang Lai, Wenjun Deng, Kan Wu, Zijie Liang
Sun Yat-Sen University
Guangzhou, Guangdong, China
This paper describes SYSU IARC team’s entire system for Mission 7a of the International Aircrafts Competition (IARC). The basic system involves an M100, an airborne computer, different kinds of sensors, communication devices and a remote control. We make full use of the optical flow, IMU data and grid pattern recognition to construct a positioning and flight control system. The vision component processes the cameras input to identify and track the ground robots, generating an effective path in real time while avoiding obstacles with threat avoidance system. To realize top-down control, we utilize simulation module to determine a practical strategy for chasing the targets. Besides, the safety measures and experiments are also discussed in detail.
Team Technical Paper
Peng Xinyu, Dong Yuan, Xu Jiangping, Lu Geng
Tsinghua University
Beijing, China
This paper describes our aerial robot, including the mechanical system, hardware system and software system. This robot is designed to be an autonomous robot for robot herding tasks, capable of navigating using visual information and finishing different tasks, mainly following the rules of IARC 2016. This paper introduces both the hardware design of the robot and the algorithms we have proposed and implemented.
Autonomous Quadcopter for Multiple Robot Tracking and Interaction in GPS-Denied Environments
George Lachow, Alexander Khoury, Jason Quach, Eric Ho
University of California, San Diego
La Jolla, California, United States
This paper describes the entire Unmanned Aerial Vehicle designed by IEEE at UC San Diego to compete in the International Robotics Competition. The system capable of interacting with objects, avoiding obstacles, within a GPS-Denied environment autonomously. Sensing the environment is strictly performed through a downward facing camera with a variety of sensors in order to establish positional estimates about the aircraft itself and the external agents. Immediately following these estimates an appropriate reaction is incurred. A majority of our system testing was performed in a software in loop simulator, enabling us to feasibly assess and evaluate a multitude of situations.
Autonomous Aerial Robot Localization and Target Detection and Manipulation in a GPS-denied Environment
Alex E. Bennett, Alexander A. Rickert, William B. Crawford, Elijah D. Lamppin, Devon D. Warman
University of Louisville
Louisville, Kentucky, United States
In this paper, a solution for Mission 7a of the 2017 International Aerial Robotics Competition is presented by Redbird Robotics of the University of Louisville. This competition poses a task that requires an autonomous aerial robot, without GPS, to herd 10 ground robots into a corral without colliding into moving obstacles or leaving the confines of the arena. Redbird Robotics’ solution is to build a single quadcopter that utilizes multiple onboard cameras for ground robot localization, an optical flow module, and a 1-dimensional LIDAR unit for accurate position estimation. The flight of the quadcopter is controlled through an onboard NVIDIA Jetson TX2 running ROS. This computer handles communication with the chosen Pixhawk flight controller running the PX4 flight stack through MAVROS and the MAVLink protocol. Real-time monitoring of the quadcopter is maintained from a ground station running a ROS node that communicates with the aerial robot over a Wi-Fi network.
Autonomous Quadrotor for the 2017 International Aerial Robotics Competition
Nicholas Eckardt, Sasawat Prankprakma, Nigel Swenson, Cheng Jiang, Steven Schulte, Sajan Patel
Universiy of Michigan
Ann Arbor, Michigan, United States
Unmanned Aerial Vehicles (UAV) are becoming more popular for both professional and casual uses, but are restricted to open areas and require GPS for navigation. Vehicles capable of flying in environments without relying on GPS will pave the way toward redefining currently outdated and expensive methods of structural inspection, search and rescue, and law enforcement operations that often take place in areas with limited GPS availability. Michigan Autonomous Aerial Vehicles (MAAV) designs and builds lightweight quadrotor UAVs capable of stable, autonomous flight without GPS. MAAV’s vehicle will compete in the 2017 International Aerial Robotics Competition (IARC) where it will demonstrate its ability to autonomously manage a herd of ground vehicles in an open environment. Using a combination of control, computer vision, and path planning algorithms, it will herd ground robots over the goal line in the required time.
University of Nebraska-Lincoln Unmanned Aerial Vehicle Design Team
Daric Teske, Nick Johs, Elliot Sandfort
University of Nebraska-Lincoln
Lincoln, Nebraska, United States
The University of Nebraska-Lincoln International Aerial Robotics Design Team is creating an autonomous quadcopter platform that will allow for easier future development, with the intention of completing the mission in the coming years. The platform is a stack of hardware, embedded flight control software, and image processing and object detection software. Combining many different backgrounds, this design team is composed entirely of undergraduate students, with majors including computer science, computer engineering, mechanical engineering, and electrical engineering.
Aerial Robot Design for Ground Robot Interaction and Navigation without Landmarks
Aaron Miller, Levi Burner
University of Pittsburgh
Pittsburgh, Pennsylvania, United States
A design for an autonomous drone's mechanical, electrical, and software system capable of finishing the International Aerial Robotics Competitions Mission 7a will be presented. The solution focuses on the ability to herd a group of ten ground robots across a single side of an arena while avoiding tall cylindrical obstacles. The sensors used for orientation and position estimates will be described as well as their intended purposes and why they were selected. In addition, a monitoring system will be described that enables safe recovery of the drone during software failure. The drone is of a quadrotor design, 107cm across and weighs over 4kg. Five cameras, a planar lidar, two ground distance sensors, ground contact switches, an accelerometer, a gyroscope, and a magnetometer are used for state estimation and target tracking. A real time flight controller is used for active stabilization, while a Jetson TX2 and a ground station are used for image processing, high level control, and planning.
IARC Technical Paper
Samid Ahmed, Nicholas Boeker, Eric Johnson, Aaron Karns, Mark Loveland, Umer Salman
University of Texas at Austin
Austin, Texas, United States
Texas Aerial Robotics will compete in Mission 7 of the International Aerial Robotics Competition (IARC) in 2017. Our goal is to direct Roombas (iRobot Create 2's), known as ground robots, using a quadcopter equipped with a Logitech C210 camera for vision, a Nvidia Jetson TX1 for vision processing, and a Pixhawk for controlling the quadcopter. The mission takes place in a GPS denied environment, so gridlines will be used as reference positions.
ZMART Technical Report
No authors listed
Zhejiang University
Hangzhou, Zhejiang, China
The Zhejiang University Micro-Aerial Robotics Team (ZMART) has prepared to participate the 2017 International Aerial Robotics Competition (IARC). Our team aims to demonstrate interaction with one moving object while autonomously navigating in a sterile open environment. The basic system architecture consists of a quadrotor helicopter platform, control units, different kinds of sensors, communication module and a base station. The hardware structure, as well as the algorithm structure, will be introduced in this report.
© Copyright 2017 International Aerial Robotics Competition