Aerial manipulation—A literature survey
Introduction
In recent years, a significant growth in Unmanned Aerial Vehicle (UAV) industry has been realized. As an example, in the United States only, there were approximately a million UAV or “drone” gifts for Christmas 2015 [1]. To this date, UAVs have been used in applications such as remote sensing of agricultural products [2], forest fire monitoring [3], search and rescue [4], border monitoring [5], transmission line inspection [6], and plant assets inspection [7]. Fully functional UAVs for plant inspection have appeared as recently as 2010 for UK onshore oil refineries [8]. In 2012, the supermajor oil and gas company, British Petroleum, established research teams to develop the necessary technologies to use UAVs for oil pipeline inspection in Prudhoe, Alaska [9] and over the course of only a few years, the technology has matured to become the standard practice for onshore and offshore platforms [7]. The above achievements have benefited various industries tremendously; however, an important common shortcoming in the mentioned applications is that the UAV is employed to merely sense, monitor and “see” the environment, but physical interaction with the environment is strictly avoided. Motivated by this, researchers in the last few years have begun examining applications in which a UAV is required to perform perching, grasping, and manipulation [[10], [11], [12], [13], [14], [15]]. This new area of research, usually known as aerial manipulation, encourages physical interaction of the UAV with its surrounding environment and enables UAVs to perform a whole new set of missions.
Aerial manipulation falls within a well-studied broad research category known as mobile manipulation. However, most of the research carried out in mobile manipulation focuses on ground robots. The main distinct challenges in the aerial manipulation problem are:
- 1.
Unlike ground robots, UAVs do not have a stable base and therefore forces and torques generated by the presence and movement of the manipulation mechanism and/or the payload directly affect the vehicle’s position, attitude and even its stability;
- 2.
Unlike ground robots, the performance of UAVs’ propulsion system vary in close vicinity of the ground and/or walls;
- 3.
UAVs are often underactuated platforms with highly nonlinear coupled dynamics, introducing further complications into their control design; and
- 4.
UAVs usually have stringent payload weight constraints and therefore cannot accommodate industrial dexterous robotic manipulators.
The above challenges encourage the development of a new research theme for the aerial manipulation problem.
An aerial manipulation system, viz. Unmanned Aerial Manipulator (UAM) hereafter, consists of two subsystems, namely a UAV and an interaction/manipulation mechanism (such as a robotic manipulator or a rigid tool) employed to physically interact with the environment. A rich amount of research literature and a number of review papers have been published on either of the above subsystems. As an example, in [16], a comprehensive survey of control algorithms for UAVs was presented. In that work, a number of schemes such as Proportional–Derivative–Integral (PID), Linear Quadratic Regulator (LQR), H, sliding mode variable structure, backstepping, and adaptive control along with their advantages and drawbacks in the control of UAVs with Vertical Take-Off and Landing (VTOL) capabilities, e.g. the quadcopter, were discussed. A detailed review of motion planning and trajectory planning algorithms for UAV guidance was also presented in [17]. Later, a review of path planning algorithms in the presence of disturbances and uncertainty was given in [18]. Also, a comprehensive literature survey on manipulation and grasping in robotic manipulation was given in [19]. While the above works summarize a broad body of literature on UAVs and robotic manipulators, they do not specifically discuss UAMs. In fact, to the best of authors’ knowledge, there is no published review paper on aerial manipulation including mission scenarios, mathematical modeling, and control schemes used in UAMs.
Aerial manipulation is a new field of research. Some of the pioneering works in this area appeared in the first years of the current decade [[10], [11], [20], [21], [22]] where the manipulation usually consisted of a gripper rigidly attached to a UAV body or was based on tethered configurations. Over the course of a few years, aerial manipulation has considerably evolved and more recent works, e.g. [[14], [15], [23], [24], [25], [26]], address challenging problems such as valve turning (see Fig. 1) and pick-and-place by several Degrees-of- Freedom (DoF) robotic manipulators. The authors believe that the coming years will bring further advancement in aerial manipulation and will enable more practical and reliable UAMs in a variety of applications.
The rest of this paper is organized as follows. Section 2 describes the most commonly used UAVs and manipulation/interaction mechanisms in aerial manipulation systems. Section 3 thoroughly studies missions and scenarios realized, to this date, in the area of aerial manipulation. Two main modeling methodologies and various control schemes in aerial manipulation are presented in Sections 4 Modeling of a UAM, 5 Estimation and control of UAM, respectively. Conclusions and directions for future research are presented in Section 6.
Section snippets
The physical subsystems of UAMs
In general, an aerial manipulation system contains two main physical subsystems, a UAV platform and a manipulation mechanism, with the necessary sensors and control systems for its autonomous or semi-autonomous functionality. In this section, we describe the most common subsystems of a UAM as well as the possible sensory configurations considered for various applications.
Missions and operational scenarios
In this section, the missions and operational scenarios considered for UAMs are discussed.
Modeling of a UAM
In this section, kinematic and dynamic modeling of a UAM and its interaction with the environment are discussed. For the sake of simplicity, a UAM consisting of a quadcopter and a manipulator with DoF is considered. The UAM model presented in this section is similar to that in [23] and can be used for other aerial platforms (such as tri-rotors or ducted-fan vehicles) with minor modification. Consider the schematic view of a UAM in Fig. 7, where is an inertial coordinate frame, a frame
State and parameter estimation
Accurate and efficient estimation of the state and parameters of UAMs is an important element of autonomous aerial manipulation missions. However, dynamic modeling of a UAM results in coupled and highly nonlinear equations of motion and therefore conventional estimation algorithms are generally difficult to apply for UAM applications. Among the limited works in this context, a least square problem was formulated in [11], using control input and acceleration measurements, to estimate the mass,
Discussion and future directions
The area of aerial manipulation was studied in this paper. The findings are summarized as follows:
-
Various types of UAV platforms have been used in aerial manipulation applications. Strong hovering capability is the main characteristic that such vehicles should possess and therefore fixed-wing UAVs are not appropriate platforms for aerial manipulation applications. Among vehicles with hovering capability, quadcopters are the most popular ones, followed by helicopters. Other
Acknowledgment
This research was financially sponsored through a Discovery Grant # 203060-12 from the Natural Sciences and Engineering Research Council of Canada, NSERC.
Hossein Bonyan Khamseh received his B.Sc. and M.Sc. in aerospace engineering from Amirkabir University and Shahid Beheshti University in 2008 and 2010, respectively. He is currently pursuing his Ph.D. at Mechanical and Industrial Engineering Department of Ryerson University, Toronto, Canada. His current research interests include aerial manipulation, aerial robotics, estimation, control, and visual servoing. To this date, He has authored and co-authored more than 20 conference and journal
References (165)
- et al.
Application of robotics in onshore oil and gas industry—A review Part I
Robot. Auton. Syst.
(2016) - et al.
Modelling and control of a flying robot interacting with the environment
Automatica
(2011) - et al.
Modeling and control of aerial manipulation vehicle with visual sensor
IFAC Proc. Vol.
(2013) - et al.
Mechanics, control and internal dynamics of quadrotor tool operation
Automatica
(2015) - FAA Nightmare: A Million Christmas Drones, 2016. [Online]. Available:...
- F.G. Costa, J. Ueyama, T. Braun, G. Pessin, F.S. Osório, P.A. Vargas, The use of unmanned aerial vehicles and wireless...
- et al.
Cooperative forest fire surveillance using a team of small unmanned air vehicles
Int. J. Syst. Sci.
(2006) - H.A.F. Almurib, P.T. Nathan, T.N. Kumar, Control and path planning of quadrotor aerial vehicles for search and rescue,...
- et al.
Decentralized cooperative aerial surveillance using fixed-wing miniature UAVs
Proc. IEEE
(2006) - H. Li, B. Wang, L. Liu, G. Tian, T. Zheng, J. Zhang, The design and application of SmartCopter: An unmanned helicopter...
Cooperative manipulation and transportation with aerial robots
Auton. Robots
Hybrid adaptive control for aerial manipulation
J. Intell. Rob. Syst.
Introduction to feedback control of underactuated VTOLvehicles: A review of basic control design ideas and principles
IEEE Control Syst.
A survey of motion planning algorithms from the perspective of autonomous UAV guidance
J. Intell. Robot. Syst.
Survey of motion planning literature in the presence of uncertainty: Considerations for UAV guidance
J. Intell. Robot. Syst.
Planning and control for cooperative manipulation and transportation with aerial robots
Int. J. Robot. Res.
Economics of transmission-line maintenance and repair by helicopter
Trans. Am. Inst. Electr. Eng. Part III
Maintenance of transmission lines by helicopter
Electr. Eng.
Helicopter use on fire suppression in the pacific northwest
J. For.
The situation with respect to forest entomology and pathology in Canada, 1943 to 1953
For. Chronicle
The helicopter for tropical botanical field work
Ecology
The evolution of the McDonnell model 120 crane helicopter
J. Am. Helicopter Soc.
Landing an unmanned air vehicle: Vision based motion estimation and nonlinear control
Asian J. Control
Vision-based sensor and navigation system for autonomous air refueling
J. Guid. Control Dyn.
Nonlinear trajectory control of multi-body aerial manipulators
J. Intell. Robot. Syst.
Cited by (0)
Hossein Bonyan Khamseh received his B.Sc. and M.Sc. in aerospace engineering from Amirkabir University and Shahid Beheshti University in 2008 and 2010, respectively. He is currently pursuing his Ph.D. at Mechanical and Industrial Engineering Department of Ryerson University, Toronto, Canada. His current research interests include aerial manipulation, aerial robotics, estimation, control, and visual servoing. To this date, He has authored and co-authored more than 20 conference and journal papers and a book chapter. He has also served a number of roles including conference/journal reviewer and co-chair, guest editor, invited speaker, and instructor in various capacities.
Farrokh Janabi-Sharifi (S’91–M’95–SM’02) is a Professor of Mechanical–Industrial Engineering and the Director of Robotics, Mechatronics and Manufacturing Automation Laboratory (RMAL) at Ryerson University, Toronto, Canada. He received the Ph.D. degree in Electrical and Computer Engineering from the University of Waterloo, Canada in 1995. He was NSERC postdoctoral fellow and instructor in the Center for Intelligent Machines and Department of Electrical–Computer Engineering of McGill University (1995–1997). His research interests span over optomechatronic systems with the focus on image-guided control and planning of robots. He has been a visiting professor in KAIST, Taejon, Korea; IRISA–INRIA, Rennes, France; and TUM, Munich, Germany. He has also been Organizer and/or Co-organizer of several international conferences on optomechatronic systems control. He was General Chair of 2010 International Symposium on Optomechatronic Technologies, Toronto, Canada. He currently serves as the Associate Editor of several journals including International Journal of Optomechatronics and technical editor of IEEE/ASME Transactions on Mechatronics.
Abdelkader Abdessameud is an Assistant Professor at the school of science, engineering and technology (SSET), Pennsylvania State University, Harrisburg. He received a Ph.D. degree in Electrical and Computer Engineering (Robotics and Control) at the University of Western Ontario (UWO), Canada, in 2010. Prior joining Penn State Harrisburg, he held several post-doctoral positions at the UWO and Ryerson University, Toronto. From 2001 to 2007, he was affiliated to the University of Boumerdes, Algeria, as a full-time Instructor (Maître Assistant Chargé de Cours) and Research Assistant. His research interests lie in the broad area of robotics and control with applications to unmanned aerial vehicles, aerial manipulation, attitude estimation and control, and the cooperative control of multi-agent systems.