IEEE Robotics & Automation Magazine - December 2018 - 77

Cooperative Aerial Manipulation
Flying robots such as multirotors are now able to monitor a
local spot, take aerial photographs, and create a high-resolution topographic map of a previously unknown space. In
these applications, the robots do not need to interact with
external environments physically, so the potential remains
for aerial robots to perform more interesting tasks by fully
using their maneuverability in three-dimensional space.
Growing interest in new aerial applications has promoted
research on interactive tasks, such as transporting an object
and performing specific tasks by applying forces on a surface. Interactive tasks accomplished by aerial robots are
often referred to as aerial manipulation, and the aerial
robotic platforms equipped with additional tools or devices
to perform such tasks are known as aerial manipulators.
The Needs and Challenges of Aerial Cooperation
Aerial robots may encounter tasks that require a greater
amount of payload than that allowed for a single robot. Aerial
manipulators can benefit from depending on colleagues, just
as humans do-for example, by cooperatively holding the
common payload and distributing the weight (Figure 1). Such
aerial cooperation is inevitable for expanding the application
of aerial manipulation because such cooperation enables a
much wider range of tasks.
However, cooperation is not easy for aerial vehicles. Cooperative aerial manipulation requires a complex high-dimensional system involving the state variables of individual
multirotors, manipulators, and an object. Moreover, multiple
kinematic constraints arise between the connection of the
platforms. Especially in the air, such constraints may cause
dangerous situations for aerial vehicles. When multiple aerial
manipulators jointly hold one object, the agents experience
inherent physical interaction with the colleague(s), even during simple hovering. Moreover, the shape or size of the object
can restrict the maneuverability of aerial vehicles. Specifically,
in an environment cluttered with obstacles, the available

Figure 1. Aerial manipulators cooperatively moving a common
object.

configurations of aerial manipulators are further restricted,
and each robot should move in ways that take account of
counterpart robot(s) and the object.
Consequently, to maximize the benefits of collaboration,
two capabilities are essential:
● First, each aerial manipulator must be well controlled to
minimize the influence caused by the interaction between
agents, disturbances, or the uncertainty associated with
specific tasks.
● Secondly, when environmental constraints are observed by
onboard sensors such as camera and lidar, a safe trajectory
compatible with complex dynamics must be computed
quickly enough to deal with such constraints swiftly.
The next section summarizes the existing research efforts on
these crucial aspects.
Related Works
Several research groups have developed cooperative aerial
systems with various manipulation parts, as illustrated in Figure 2. Multirotors, the most straightforward gripping mechanism, were rigidly attached to an object using a gripper in [1].
The attached object and multirotors composed a new rigid
body system. The desired net force and torque required to

Multirotor 1
State Variables

Multirotor 2
State Variables

Kinematic Constraints
Manipulator 1
State Variables

Kinematic Constraints

Manipulator 2
State Variables
Kinematic Constraints

Common Object
State Variables

Figure 2. A cooperative aerial manipulation system with a common object.

december 2018

IEEE ROBOTICS & AUTOMATION MAGAZINE

IEEE Robotics & Automation Magazine - December 2018

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2018