Supplementary Information: Multi-Modal Mobility
Morphobot (M4) with Appendage Repurposing for
Locomotion Plasticity Enhancement
Eric Sihite
ଵ
, Arash Kalantari
ଶ
, Reza Nemovi
ଵ
, Alireza Ramezani
ଷ
∗
, and Morteza
Gharib
ଵ
ଵ
Aerospace Engineering Department, California Institute of Technology, 1200 E California Blvd,
Pasadena, California, USA.
ଶ
Jet Propulsion Laboratory (JPL), 4800 Oak Grove Drive, M/S 82-105, Pasadena, California, USA
ଷ
Electrical and Computer Engineering Department, Northeastern University, 360 Huntington Ave,
Boston, Massachusetts, USA
∗
Corresponding author email:
a.ramezani@northeastern.edu
Supplementary Notes
This supplementary note includes three
composite figures referenced in the main dra
Ō
. These figures
illustrate:
1-
A conceptual design of a machine with extensive locomo
Ɵ
on plas
Ɵ
city
(Supplementary Figure 1).
2-
Simulated mul
Ɵ
-
modal path planning results based on mul
Ɵ
-modal PRM (Supplementary Figure
2).
3-
WAIR simula
Ɵ
on trajectory (Supplementary Figure
3).
Supplementary Figure 1:
Shows an imaginary morpho-func
Ɵ
onal robot, where body morphing and appendage repurposing are
u
Ɵ
lized to achieve several modes. This concept is not physically realizable due to large range of mo
Ɵ
ons in joints, part collisions,
and actua
Ɵ
on challenges. It is only used for illustra
Ɵ
on purposes. Note that within each numbered state, other modes of
opera
Ɵ
on are conceivable (e.g., walking, tro
ƫ
ng, galloping, etc., within mode #4), yielding addi
Ɵ
onal locomo
Ɵ
on plas
Ɵ
city.
Supplementary Figure 2:
Illustra
Ɵ
on of the path planning algorithm for naviga
Ɵ
ng an environment using both ground and aerial
mobility.
(a)
Representa
Ɵ
on of the set of nodes generated by a uniform grid discre
Ɵ
za
Ɵ
on.
(b)
Example of graph generated by
the 3D MM-PRM Algorithm with the following parameters:
R
= 4 meters,
Nw
= 300, and
Nf
= 300. The MM-RPM method
generates a significantly reduced amount of nodes which greatly reduces the computa
Ɵ
onal
Ɵ
me and cost of performing the
path-finding algorithm.
(c)
The trajectories generated by the 3D A
∗
path planning algorithm on three different environments.
The red box represents an obstacle that cannot be driven over, while the green box represents a drive-able pla
ƞ
orm.
Supplementary Figure 3:
Plots in the first two rows show the pitch angle states and controller inputs during the thruster-
assisted self-uprigh
Ɵ
ng maneuver. Plots in rows 3, 4, and 5 depicts the states and controller inputs during WAIR. The non-zero
thruster force on the slope indicates that the thrusters assisted in driving up the incline.