IEEE Robotics & Automation Magazine - June 2017 - 77

capsule motion, interfering with visual diagnostics. In contrast,
pneumatic pressurization does not rely on reciprocating parts
and can provide an inherently stable delivery pressure to
enhance capsule stability. Another notable advantage of pneumatic pressurization is the independence from electrical
power. Because compressed air can be carried in commercial
tanks, this approach offers unique advantages in terms of
portability and utility in LMICs.
Experimental Analysis
Force Characterization
Characterization of water actuation force was performed to
establish the relationship between valve position and jet force
and to examine hysteresis in jet control. Jet force was measured using a calibrated load cell (ATI Industrial Automation,
Apex, North Carolina, model NANO17; resolution 0.318 g
force). The capsule was connected to the load cell using a
265-mm rod, and jet force was measured using a cantilever
arrangement [Figure 4(a)]. Five trials were conducted showing good repeatability of results. At a standard system pressure of 80 lbf/in2, the maximum measured actuation force
was 0.128 N with the valve fully opened. The measured jet
force as a function of valve position [Figure 4(c)] exhibits a
linear region in the center of the input range, which is favorable for capsule controllability.
Although control is repeatable, hysteresis is present
between the opening (unloading) and closing (loading) of the
valve. This discrepancy is likely due to positional inaccuracies
in the pinch valves themselves, rather
than a fluid dynamical hysteresis.

Range of Motion Using a Single Jet
This experimental trial aimed at understanding the controllability of the capsule while throttling a single jet from fully
closed to fully open. Camera stability for internal visualization
is the main requirement for any endoscopic platform. This
trial was carried on to quantify the number of stable positions
the capsule can reach and the maximum displacement that
can be obtained with respect to the free length of the tether.
To be considered a stable position, the capsule must be still
enough to use the camera for visual inspection. The capsule
motion was monitored using a 6-DoF magnetic coil (0.9-mm
diameter, 12-mm length) embedded in the capsule and excited with an electromagnetic transmitter (Northern Digital
Inc., Waterloo, Ontario, Canada, model AA138). During the
study, the tether was secured and held vertical using an aluminum metallic arm with a custom 3-D printed holder to provide a well-defined pivot point [Figure 5(a)]. Water was fed to
the capsule internal nozzles using the multilumen catheter
described in the "Multilumen Tether" section. Three different
tether lengths (L) of 6, 9, and 12 cm were tested to obtain the
relationship between lateral and vertical displacement and
therefore quantify the maximum motion with respect to the
vertical position [Figure 5(a)]. A single sweep motion [Figure 5(b)] was programmed using the suitcase control electronics. The sweep consisted of gradually controlling the jet

Flow Rate (mL/min)

500
Force/Torque
Sensor

Unloading
Loading

400
300
200
100
0

10
20
Valve Position

(b)
Metal Rod

0 20
(mm)
(cm)
0 2

Nozzle

Jet Force (mN)

Flow Rate Characterization
Jet flow rate was measured during jet
force testing using an ultrasonic flowmeter (Atrato, Sherborne, United Kingdom, model Titan 760), which provides
the instantaneous flow rate through
the jets. Basic fluid dynamics theory
dictates that for a nondeforming system at steady state, jet force is a function of flow rate alone. This relationship
provides a basis for control of the jet
actuation force.
As expected from the jet force measurements, hysteresis in flow control is
present in the experimental data [Figure 4(b)]. Although fluid flow should
show no hysteresis, pinch valves rely
on a mechanical drivetrain and show
some error in control. The hysteresis is
seen to increase as the valve clamping
force increases, due to the greater forces imposed on the valve drivetrain.
When loaded, both frictional forces
and motor dynamics contribute to hys-

teresis in the drivetrain. Using a fixed upstream pressure of 80
lbf/in2, the maximum measured flow rate was 410 mL/min,
which agrees with classical fluid modeling equations.

140
120
100
80
60
40
20
0

Unloading
Loading

10
20
Valve Position

(a)

(c)

Figure 4. (a) The experimental setup during the force characterization experiment. (b) The
jet flow rate as a function of valve position. (c) The jet actuation force as a function of
valve position.

June 2017

IEEE ROBOTICS & AUTOMATION MAGAZINE

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