IEEE Robotics & Automation Magazine - December 2017 - 97

Table 4. The task description, main rules, required skills, and scores of
the six tasks composing the C-PAPR. (Photos/images courtesy of ETH Zürich.)
Puzzle

Scheme

Task Description
●

Shelf and Tray

Wire Loop

●

●
●

Breakfast

●

●
●

Carry

Hang Up

●

Transfer a 3 × 3 grid of
square wooden bases
from one puzzle frame to
another.
Handles have different
shapes.
Move a wire loop from one
end of a metal wire course
to another.
The wire loop is conductive.

Transfer items from shelves
and drawers onto a tray.
Carry the tray over a ramp.

Main Rules
●

●

Open a door and continue
down a ramp.
Reach a table and screw
a light bulb into a table
lamp.
Open a water bottle and a
jar.
Unwrap a sugar cube.
Cut a loaf of bread.
Use a can opener.
Pin a T-shirt onto a metal
line by manipulating pins.
Close (buttons and a zipper) and hang two jackets
using the hangers.
Carry bags, parcels, and
balls over flat ground.
Carry the items up and
down stairs and place
them on a table.

positions: hanging freely below the line (pointing down
and able to swing) [Figure 7(s)] or stably placed on the line
(pointing up) [Figure 7(t)]. In both orientations, he found
the most stable and easy-to-use grasp to be between the
pinky and ring fingers and the palmar surface without
needing to adjust the wrist orientation (assuming a neutral
wrist starting orientation).
Carry
Our pilot had practiced stacking all items onto the large box
and lifting the stack by the box handles using his natural and
prosthetic hands [see Figure 7(u)-(x)]. During the final race,
however, he chose to make two trips, slightly increasing his
time on this task.

●

Required Skills

The pieces could only
be lifted by the handle.
The handle could only
be manipulated using
the prosthetic terminal
device.
Any contact with the
wire course resulted in a
task failure.
The start and finish were
safe zones.

The bowl, plate, coffee
cup, cutlery, and
light bulb could only
be handled with the
prosthetic device.

The task could be
completed using either
hand.

●
●
●

●

●
●
●

●

●
●
●
●

●

The clothespin could
only be manipulated
using the prosthetic
terminal device.
The pilot could make as
many trips as desired.
Objects varied in weight
from 0.4 to 5 kg.

●
●

●
●
●

Points

Adaptivity
Grip force
Manipulation

115

Stability of the control
signals
Precision
Compensation ability
Wrist, elbow,
shoulder, and trunk
coordination

102

Adaptivity in real
environment
Ability to use the
prosthesis during
movements

130

Bimanual tasks
Power
Compensation ability
Wrist

104

Bimanual task
Precision

108

Force
Control signal
Robustness

101

C-PAPR Results
Total scores and time (in seconds) for each task and each
team admitted in the final race of the C-PAPR are reported
in Table 5. The results are split by prosthetic hardware type
into BP (one team), single-grasp powered (one team), and
multigrasp powered, regardless of control mode (five
teams), and compared to results of the SoftHand Pro team.
A breakdown of the time to task completion of the teams
in the finals can be found in Figure 8(a). However, in cases
where more than one team used similar prosthetic hardware (as was the case for multigrasp-powered prostheses),
the mean plus-or-minus-one standard deviation is presented. As shown in Figure 8 and Table 5, the SoftHand
Pro team's results were highly competitive. Our pilot
DECEMBER 2017

IEEE ROBOTICS & AUTOMATION MAGAZINE

Table of Contents for the Digital Edition of IEEE Robotics & Automation Magazine - December 2017