IEEE Robotics & Automation Magazine - December 2018 - 43

Discussion and Conclusions
Throughout the development of the Voliro
system, a range of issues were encountered
on which we are basing our development of
future platforms. While building the tilt
motors into the arm tips has a range of advantages, it also has the following practical
limitations:
1) Significant backlash occurs as the parts
wear due to the difficulty of maintaining
(a)
a tight fitting.
2) Accessing and replacing the internal
motors is challenging.
3) The limited airflow can cause overheating issues.
4) The overall system is difficult to repair,
as the forces from a crash can often be
transmitted down the arms to the internal components.
5) The thrust-to-weight ratio of the system
is marginal when hovering in its least
efficient configurations, limiting its ability to carry payloads.
In current follow-up designs, these issues
(b)
are being addressed by the design of arms
that employ dual-thrust motors and are
driven by servos in the base through a simple clutch. On the software side, as previously mentioned, some problems can be caused
by slow rotor-tilting dynamics when the system hovers at 90c.
Overall, however, we demonstrated how
the basic idea of a hexacopter with tiltable
rotors can be used to achieve omniorienta(c)
tional maneuverability. At the same time,
we avoid wasting energy on the generation
Figure 12. The Voliro UAV during its wall interaction. (a) A vertical orientation is
of counteracting forces, which is an issue
achieved to align the three-sphere module with the wall. (b) The wall is approached
while maintaining the aforementioned orientation. (c) The platform is able to
that inhibits the use of fixed-orientation
establish stable contact with the wall and move in any direction while maintaining
rotors in omnidirectional designs. We preforce against the wall.
sented the mechanical design of the platform and a compact tiltable rotor. The
such that the position and the orientation were still slightly employed control allocation technique has the advantage of
coupled. This can be partially explained by the slower limiting the computational effort required to compute tilting
dynamics of the tilting motors when compared to the angles and rotor speeds.
thrust motors. The experiments showed that the roll and
In various experiments, we demonstrated a transition
pitch angles were tracked better than the position. This from horizontal to upside-down flight and physical interacoccurred because the rotation dynamics are controlled tion with a wall. While the increased maneuverability gives
mainly by changing the rotational speeds of the rotors and rise to a broader scope of applications, some of them may
using their fast dynamics. Deviations in position and yaw require a powering tether to overcome the limitations
are generally corrected by the tilting of the rotors and are imposed by battery life. In conclusion, the tiltable-rotor mulmore sensitive to their slow dynamics.
tirotor showed great capabilities and can push the boundaries
While flying at the pitch angle of 90c, the allocation of what is currently achievable by standard multirotors.
demanded inadmissible rotor speeds, leading to two thrust
motors becoming saturated. This is further discussed in Acknowledgments
the "Control Allocation" section. Nevertheless, the system We would like to acknowledge Philipp Andermatt, Cliff Li,
was still able to track the desired position and orientation.
Alexis Müller, Kamil Ritz, and Kevin Schneider, who were
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