IEEE Robotics & Automation Magazine - June 2018 - 36

0.71 m

y
z

Force (N)

1
0.27 m

0.33 m

0
-1

x-Axis
y-Axis
z-Axis

-2

Figure 10. A holed board used for the peg-in-hole experiment.

4
5
6
Time (s)

Average Force (N)

Force (N)

20
0
-20
-40

x-Axis

y-Axis

25
20
15
10
5
0

-60
10
15
Time (s)

Figure 12. The measured forces for the peg-in-hole experiment
with the proposed LIP-HIA mechanism.

0.64 m

4
x

z-Axis

20
25
Time (s)

Impedance Control
Admittance Control

Impedance Mean
Admittance Mean

Figure 11. The measured forces for the peg-in-hole experiment
with only the HIA mechanism and admittance control.

Figure 13. The average force and time needed to execute the
task for the 100 trials performed by the ten volunteers.

mechanism. A comparison between the macro-mini HIA-LIP
mechanism and a state-of-the-art admittance controller commonly used in pHRI [5] was conducted by performing a pegin-hole task. The peg and the holed board used in [14] were
used for this experiment. Figure 10 shows the holed board
used and the sequence that the user needed to follow. First, the
peg started at the center of the board (position 0). The user
then moved the peg into the first hole (position 1), inserted it
to a depth of ~2.5 cm, and then withdrew it. Afterward, the
peg was moved to the second hole, the third hole, and then to
the last. After the peg was withdrawn from the last hole, it was
returned to its original position at the center of the board.
A group of ten volunteers, all males between the ages of 23
and 52, participated in the peg-in-hole experiment. First, the
subjects were asked to perform five trials of the peg-in-hole
sequence as fast as possible using only the HIA mechanism
with a force sensor mounted on its end effector. In this case,
the forces measured by the sensor at the base of the peg were
used to move the HIA robot using an admittance controller.
The admittance parameters used for this task were found heuristically to generate the fastest motions possible without producing unstable contacts (m = 20 kg and c = 300 N/m) .
During the experiment, the time taken to complete the
task was measured, as well as the interaction force between

the user and the handle. The measured forces obtained from a
typical trial are shown in Figure 11, and the peg insertion
positions are highlighted at the top of the graph for reference.
The forces reached values of approximately 60 N and the time
of execution was just under 25 s.
The subjects were then asked to perform five trials of the
same peg-in-hole sequence using the macro-mini HIA-LIP
mechanism. The handle of the LIP mini mechanism shown in
Figure 2(a) was replaced with the peg that served as the handle. The peg was mounted horizontally to perform the task,
and a force sensor was attached to the base of the handle to
measure the forces (in all three Cartesian directions) applied
by the user on the handle when moving the mini LIP mechanism. The force measurements were not used in the controller.
The measured forces obtained from a typical trial are
shown in Figure 12. In this case, the maximum force is approximately 2.3 N, which represents ~4% of the maximum
forces exerted with the admittance controller. The time of execution is just above 9 s, which is significantly faster than with
the admittance controller.
For each trial, the average value of the magnitude of the
interaction force was computed. The average force and the time
required to execute the task for each trial with the admittance
control and the LIP-HIA control are indicated in Figure 13. It is

IEEE ROBOTICS & AUTOMATION MAGAZINE

JUNE 2018

IEEE Robotics & Automation Magazine - June 2018

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