IEEE Robotics & Automation Magazine - December 2018 - 15

so the arm movement does not modify the manipulator's
CoM. Additionally, any contact forces from the interaction
with the environment are transmitted directly to the CoM
without generating perturbation torques.
Manipulation more advanced than simple grasping or
force exertion over a surface requires the use of aerial manip-
ulators with dual arms. Very few dual-arm aerial manipula-
tors have been developed (as, for example, [15]) that use two
3-DoF arms for valve turning. But in AEROARMS, two new
anthropomorphic dual-arm aerial manipulators with four
and five joints (see Table 1) have been developed, because
this configuration maximizes dual-arm manipulability. The
first is an octorotor with arms weighing 0.9 kg. The maxi-
mum load that can be lifted by the two arms with stiff joints is
0.7 kg [Figure 1(c)].
Then, taking into account that collisions and impacts can
severely affect the stability of the aerial manipulators and that
compliance has been found to be very useful [7], a dual-arm
aerial manipulator with compliant joints [16] was developed and
tested outdoors [Figure 1(d)]. It is a hexarotor, and the arms
have joints with elastic elements that can absorb the contact and
collision forces generated by physical interaction, providing a
compliant behavior that increases the system's robustness and
stability and also enables the measurement of interaction forces
for force control. The total weight for both arms is 1.3 kg, and
the maximum load that each arm can lift is 0.2 kg.
Control Systems
One of the key AEROARMS advances has been to intensively
investigate the potentiality and limitations of control architec-
tures for aerial manipulation. As far as interaction control is
concerned, a solution based on a multidirectional-thruster
platform and a 6-D flying end effector [14] represents the
successful integration of known robotic algorithms dealing
with both motion and interaction control into a platform hav-
ing a minimal sensor suite, i.e., a pose sensor plus an intertial
measurement unit (IMU), without requiring any kind of force
sensor. Stable interaction was also obtained in previous
approaches (e.g., in [3], where an aerial manipulator inter-
acted with a vertical surface). In AEROARMS, however,
experiments involving complex interaction tasks, such as peg-
in-hole with tilted holes and sliding on tilted surfaces, have
been performed.
Two main motion control approaches have been con-
sidered. The first is based on modeling the aerial manipu-
lator as a unique system and designing the control scheme
on the complete kinematic and dynamic models. See, for
instance, [17], where an integral backstepping controller
was implemented. The second approach considers the aer-
ial platform and the robotic arm as two separate entities.
Thus, two autonomous controllers are designed, one for
the aerial vehicle, where the effects of the arm on the
dynamics are handled as disturbances, and one for the
arm, where the disturbances are due to the vehicle. Such a
paradigm has been experimentally exploited, e.g., in [18],
where the aerial vehicle controller includes an estimator of

external forces and moments to compensate for the
neglected dynamics and the arm dynamics.
Vision-based techniques have also been exploited. In [19],
the feedback output from a camera attached to the end effec-
tor was adopted in a hierarchical control law. The camera
images were exploited to drive the arm end effector to a
desired position and orientation.
One AEROARMS objective is inspection through con-
tact while flying. This requires exerting forces on a surface,
thereby maintaining contact with the sensor (e.g., ultra-
sonic) installed in the
manipulator's end effec-
tor. Because interaction
In AEROARMS, however,
forces and moments can
cause large deviations in
experiments involving
motion control and/or
destabilize the system,
complex interaction tasks,
some active and/or pas-
sive compliance should
such as peg-in-hole with
be added to the vehicle
and the manipulator's
tilted holes and sliding on
end effector. Passive com-
pliance, which was ad-
tilted surfaces, have been
opted in the platform in
Figure 1(d) through the
performed.
elastic actuators of the
compliant arms, allows for
efficient coping with interaction forces and also provides
measurements of these forces that can be used to close the
feedback loop [16]. The adoption of active compliance
approaches requires measurements of the interaction
wrenches, which can be achieved, for example, by using a
wrist-mounted force/torque sensor on the arm, as in [20],
where the contact wrenches are fed to an admittance filter.
Such a control scheme has been tested on an underactuated
quadrotor equipped with a 6-DoF manipulator.
An alternative solution to the force/torque sensor is the
use of wrench estimators, as proposed in [14] and [21]. The
adoption of underactuated platforms implies that the lateral
forces in the body frame, which cannot be provided by the
aerial platform itself, must be generated through the dynam-
ic/inertial coupling between the arm and the aerial robot,
which is usually quite difficult to manage accurately. To over-
come this challenge, the admittance control paradigm was
exploited by considering the platform in Figure 1(a), charac-
terized by noncollinear, fixedly tilted propellers, which makes
it possible to independently control the translational and
angular acceleration when unconstrained and to govern any
of the six components of the exerted wrench when in contact
[14]. Case studies considered include the contact inspection
of a pipe and a peg-in-hole emulating a sensor installation.
Figure 3 shows the relevant variables for one of the experi-
ments. Videos can be found at http://homepages.laas.fr/
afranchi/robotics/?q=node/414.
Regarding the vehicle/arms coordination, a behavioral
control scheme has been proposed for handling both
december 2018

*

IEEE ROBOTICS & AUTOMATION MAGAZINE

*

15
http://homepages.laas.fr/afranchi/robotics/?q=node/414 http://homepages.laas.fr/afranchi/robotics/?q=node/414
IEEE Robotics & Automation Magazine - December 2018

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