The paper analyzes constrained tendon routing methods through analysis and simulation.
* Mathematical analysis is done in the main paper.
** The simulation code is located here.
The number of contact points (CP in the figure below) and the scalar sum of the contact force (CF in the figure below) are measured from the simulation to compare the adaptability of the robots that use position-constrained tendon routing (PTR ) and force-constrained tendon routing (FTR).
The simulation uses six objects as below.
(a) and (d) are the basic structure of the sphere and cylinder.
(b) and (d) are combined structures with the same x and y positions but different sizes and z positions.
(c) and (f) are combined structures with different sizes and x, y, and z positions.
The contact points and scalar sum of the contact force when using PTR and FTR are described below.
* We only used the basic geometry and its combined structure instead of complicated mesh structures because constraining the tendon's position made the simulation unstable. We excluded the cube shape because we found that the sharp surface of the cube makes the simulation unstable when constraining the tendon's position.
We didn't use the vector sum of the contact force since it will be near zero (in the object frame) because the object is not moving. Instead, we considered the scalar sum of the contact forces as a performance metric. A high scalar sum indicates that the object is being squeezed. This status is more effective in resisting unwanted motion, making the system more reliable against disturbances.
It consists of four spheres that have different sizes and centers.
Blue sphere: radius="5cm", position="12cm 10cm 5cm"
Green sphere: radius="4cm", position="12cm 5cm 9cm"
Magenta sphere: radius="3cm", position="15cm 10cm 14cm"
Red sphere: radius="6cm", position="17cm 15cm 13cm"
In the simulation, the orange cylinders show the contact points and the white arrows describe the contact force.