IEEE Robotics & Automation Magazine - March 2022 - 82

simulated, where participants had to grasp and transport the
object to the target location through simulated contact forces.
Gravity was enabled for manipulation, and a table was simulated
by a collision plane. Figure 3(b) and (c) depict these two
types of object interactions.
Analysis and Procedure
All results have been tested for significance (a 95% confidence
interval) via the use of a repeated-measures analysis of variance
(RM-ANOVA). A Shapiro-Wilk test verified the normality
of the data prior to the RM-ANOVA. Maulchy's test
was employed to evaluate sphericity, and in cases of violation,
a Greenhouse-Geisser procedure was implemented to
correct the DoF, assuming that
e1075 For nonparametric
..
data violating normality, an aligned-rank transform (ART)
[20] enabled the use of the RM-ANOVA on the ranked data.
A stepwise linear regression determined the effect of each
variable for conceiving our metric equation and the effect
on the predictability of the MT. The independent spatial
variables were fitted by using the criteria of the probability
of
F .005<
to enter or F .010>
nessed linear regression analysis ()r2
to remove. Finally, we harto
analyze and compare
our proposed model against existing work. Hereafter,
the significance levels are p) 1 0.05, p)) 1 0.01, and
))) p 1 0.001.
Participants
Twenty healthy participants (N = 20) were recruited (four
females and 16 males), with ages ranging from 19 to 46
(. ,. ).
27 35
nv 543== The selection criteria were that each
participant was 1) right-handed, 2) had healthy hand control,
3) had normal/corrected vision, and 4) was familiar with
video games or VR. Candidates who did not meet the criteria
were excluded. Participants were asked to find a balance
between minimizing errors and selecting the targets as quickly
as they could during selection and placing.
Pre-Exposure and Approach
Prior to commencement, participants were briefed and gave
their formal written consent, and their individual interpupillary
distance was measured for the VRHMD. Furthermore,
acclimatization to the simulation environment was allowed
prior to the experiment, via a set of 96 training exercises covering
pointing and manipulation task types, e.g., 48 for each
type. Furthermore, translational and rotational movements
were included to cover the entirety of all spatial complexities
later investigated. The training exercises were specific to the
acclimatization procedure and independent of those in the
experiment. We also randomized the order of all experiments
to counterbalance potential acclimatization and task adaption.
A total of 39,040 trials was recorded.
F(x, y, z)
F(x, y, z)
F
F
A
W(x, y, z)
W
2
W
2
ωi
Wαi′
ωi
′
αi
Fαi
(a)
Translational Task
T =
1,
0,
A
A
≤
>
(W/2)
(W/2)
i=1 (xi - yi)2.
Rotational Task
R =
1,
0,
αi ωi
α ωi
i
≤
>
, i : 3
(b)
Combined Task
C = T . R
Notations for Translational Tasks: A represents the distance between two 3D Points; hence,
A = d(Wx,y,z, Fx,y,z) = Σ 3
Notations for Rotational Tasks: αi = |Wαi
respectively. αi = αi.
′
- Fαi
|. Wαi
and Fαi represent the rotation of the target and the object in all axes (i),
Notations for Combined Tasks: Both translational and rotational requirements need to be met to be classified as success,
e.g., T = 1 and R = 1.
αi
Wαi
′
αi
Fαi
αi
(c)
Wαi
ωi
Wαi′
ωi
W(x, y, z)
A
W
2
W
2
Figure 4. The progression requirements for (a) translational, (b) rotational, and (c) combined tasks. The green cube volume illustrates
successful placements. The table shows task successes (1) and errors (0).
82 * IEEE ROBOTICS & AUTOMATION MAGAZINE * MARCH 2022
IEEE Robotics & Automation Magazine - March 2022

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - March 2022
