63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
IAC-12-A1.6.6 Page 1 of 9
USING INERTIAL MEASUREMENT UNITS FOR MEASURING SPACESUIT MOBILITY AND WORK
ENVELOPE CAPABILITY FOR INTRAVEHICULAR AND EXTRAVEHICULAR ACTIVITIES
Ryan L. Kobrick, PhD
Massachusetts Institute of Technology, USA, firstname.lastname@example.org / email@example.com
Christopher E. Carr, ScD, Forrest Meyen, Ana R. Domingues, and Prof. Dava J. Newman
Massachusetts Institute of Technology, USA, firstname.lastname@example.org, email@example.com, firstname.lastname@example.org, and
Shane E. Jacobs, PhD
David Clark Company, USA, SJacobs@davidclark.com
Human spaceflight destinations are expanding to include a multitude of environments that will offer different
mobility challenges to explorers due to varying gravity levels and surface operations. Intravehicular Activities (IVA)
suits might include a basic “get-me-down” suit for suborbital spaceflight, or a high performance pressurized pilot suit
where arm mobility and field of vision are particularly important. Future Extravehicular Activities (EVA) will likely
accommodate various spacesuit architectures including: a microgravity station/craft maintenance suit where hand
dexterity is critical; a close proximity operation suit for asteroid missions where manoeuvrability and visibility are
critical; and a planetary surface suit for the Moon or Mars where leg mobility is a key requirement. Spacesuit
kinematics are currently measured using video motion capture or photographic analysis systems. Although these
methods measure the external motion of the suit, they do not capture the physical body motions within the suit and in
the case of motion capture, they are restricted to a laboratory setting with significant overhead for camera calibration
and set-up. Inertial Measurement Units (IMUs) use accelerometers and gyroscopes to estimate relative translation
and rotation. IMU systems are mobile, low-powered, and offer an economical and efficient kinematic tracking
capability for use in a laboratory or in the field. In this study, we applied IMU sensors to study space-suited motion.
To first validate the use of IMUs for motion tracking, we tracked knee flexion angle while walking using both IMUs
and a Vicon motion-capture system, which is considered the industry gold standard for kinematic analysis. The IMU
knee joint angle average root-mean-square error with respect to the Vicon system was 5.4 ± 2.4˚, demonstrating the
potential of the new system. We then used the IMUs, in conjunction with a Contingency Hypobaric Astronaut
Protective Suit (CHAPS), to measure elbow flexion/extension, shoulder flexion/extension, and shoulder
abduction/adduction motions for unsuited, suited and unpressurized, and suited and pressurized conditions. Results
from the elbow study demonstrate our ability to capture joint angles in a laboratory environment with the goal of
being used in any environment. In general, the internal IMU angle on the subject’s body was approximately 25˚
larger than the external CHAPS IMU external angle measured. A brief discussion summarizes key findings and
identifies limitations in the test configuration. Recommendations for future implementation and testing are outlined,
and conclusions are drawn on the usability of IMUs to investigate astronaut mobility and to provide work envelope
The new human spaceflight market in suborbital
space tourism and research flights as well as new crew
capabilities to the International Space Station (ISS) will
mix customer needs with high-powered vehicles that
lack extensive flight history. It is important to monitor
passenger comfort and safety and inform future
improvements to mission elements such as spacesuits,
personal cabin space, and throttling profiles. The fast
pace of commercial orbital vehicle development will
benefit from novel mobility measurement techniques,
especially if they can be taken within the vehicles
during operational development. Inertial measurement
units (IMUs), sensors that integrate data from
orthogonal gyroscopes, accelerometers, and
magnetometers, can aid in these assessments.
The goal of this research is to develop novel
applications for IMU technology in characterization of
human motion, such as estimating orientation,
acceleration, velocity, and position during restrained or
natural movement. In particular, this work focuses on
spacesuit mobility and how IMU data can be used to
construct range of motion joint angles and eventually
work envelope definitions in a realistic test
environment. This data can aid in the development of
future space suits and improve knowledge of current
suit performance and limitations.
Systems of inertial sensors may also have many
terrestrial applications where enhanced monitoring of
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
IAC-12-A1.6.6 Page 2 of 9
human movement is beneficial. These areas include
estimation of ambulatory joint kinematics1-4, injury
rehabilitation5,6, assessment of neurological movement
disorders7, and enhancement of athletic performance8,9.
II. SPACESUIT MOBILITY TESTING
Pressure suits are worn by pilots and astronauts to
protect them from a variety of hazards including low-
pressure environments and thermal extremes. Pressure
suits worn inside the vehicle during dynamic phases of
flight, such as launch, entry, and docking are primarily
designed to protect the crewmember in the event of an
emergency. During nominal unpressurized operations,
the crewmember must be comfortable and have the
mobility to perform mission tasks, such as ingressing
the vehicle and performing flight operations. During an
emergency, the suit must enable the crewmember to
perform any operations necessary to return to safety
while protecting the crewmember from hazards. To that
end, launch and entry suits often incorporate bailout
systems, fire protection, cold-water immersion
protection and integrated flotation, which are all
dependent on the requirements and interfaces of the
vehicle10. The current set of requirements outlined by
NASA for commercial vehicles is in the ISS Crew
Transportation and Services Requirements Document
CCT-REQ-1130. It does not specifically mandate a
pressure suit, but the NASA Astronaut Office considers
Similarly, pressure suits for the emerging
commercial spaceflight industry will be primarily worn
unpressurized, but in the event of an emergency, the suit
must ensure the crewmember survives and, if necessary,
can continue to perform the necessary functions to
return to safety. It is important to note though that
pressure suit needs vary amongst the different mission
profiles, as differing levels of mobility will be required
of passengers in different vehicles. Even within a single
vehicle, the mobility requirements are varied, as pilots
must be able to continue to fly the spacecraft while
pressurized (in the event of a cabin depressurization),
while suits for passengers must simply ensure their
Understanding and quantifying vigorously how
much mobility a crewmember needs to perform each
task is critical to derive requirements that will not over
constrain the design. It is important to recognize that
increases in pressurized mobility often come at a cost,
such as an increased mass, detriment to unpressurized
comfort, and increased development costs11,12. The
mobility requirements therefore must not drive a design
beyond that which is absolutely necessary, as other
desirable characteristics of the suit may be sacrificed.
Additionally, as NASA prepares for exploration
missions outside of low earth orbit, it is increasingly
important to be able to quantify, communicate, and
validate, space suit mobility for suits worn outside the
spacecraft. These suits are always worn pressurized, and
as such pressurized mobility becomes far more critical.
One of the long term goals of space suit design is to
design suits that allow performance as close as possible
to “shirt-sleeve mobility”, such that an astronaut in a
pressurized space suit could perform all the same tasks,
with the same ease, as a geologist on earth in a t-shirt
and shorts. Research at MIT in the Man-Vehicle
Laboratory (MVL) has been moving towards this
mobility goal with incremental subsystem design of a
mechanical counterpressure BioSuitTM 13. In order to
achieve this goal, the mobility enabled by various joint
designs must be well understood and quantified.
Improvements to the joints can then in turn be
quantified, by measuring the reduction in mobility, and
understanding the physical principles responsible for the
reduction. Without continuous benchmarking and
iteration, the suit designer cannot make progress
towards a highly mobile joint. A lower cost solution for
measuring joint mobility would also help with standards
development, as several laboratories would be able to
verify the technique.
It is evident then that proper characterization of
mobility requirements – how much mobility is needed
to perform all mission tasks – as well as mobility
capabilities – how much mobility a certain space suit
enables – is absolutely essential for both government
programs and the commercial spaceflight industry.
Spacesuit Environments: IVA and EVA
There are essentially two key working environments
that must be considered for suit mobility design.
Intravehicular Activities (IVA) suits might include a
basic “get-me-down” suit for suborbital spaceflight, or a
high performance pressurized pilot suit where arm
mobility and field of vision are particularly important.
Future Extravehicular Activities (EVA) will likely
accommodate various spacesuit architectures including:
a microgravity station/craft maintenance suit where
hand dexterity is critical; a close proximity operation
suit for asteroid missions where manoeuvrability and
visibility are critical; and a planetary surface suit for the
Moon or Mars where leg mobility is a key requirement.
Because of the large amount of mass it would not be
practical to launch three or four suits on a mission, so
trade-offs must be made to condense the suit
architecture to meet the requirements of many
Several methodologies have been used to measure
mobility, though two methods have emerged as the most
common within the spacesuit community14.
Unfortunately each has its drawbacks. Photogrammetry,
63rd International Astronautical Congress, Naples, Italy. Copyright ©2012 by the International Astronautical Federation. All rights reserved.
IAC-12-A1.6.6 Page 3 of 9
the process of measuring joint angles from pictures of a
subject in the suit at the extremes of a joint’s range, has
been used to quantify pressure suit mobility dating back
at least to the Apollo program15, and through various
space suit development programs16,17 including the most
recent prototype suits developed for NASA’s project
Constellation12,18,19. This method only quantifies
isolated joint movements, making it difficult to properly
characterize the suit’s mobility for complex tasks.
Additionally, this method requires the subject to hold a
joint at “maximum” angles, which can be very workload
intensive, and as a result tends to underestimate a suit’s
full range of mobility.
The second method commonly used, developed
more recently with advances in technology, involves
three dimensional video motion capture technology.
Subjects in suits are outfitted with reflective markers,
and systems of multiple cameras are used to track the
markers as the subject performs various functional
tasks. The coordinates of the markers can be used to
measure individual joint angles using inverse kinematics
software. This method was used extensively for the
derivation of requirements for project Constellation20.
Motion capture methodology is advantageous as it
captures mobility during functional movements and
tasks, but it is costly both in terms of equipment needed
and in post-processing time. Additionally, it requires
line of sight for several (2-3 minimum) cameras on each
marker at all times, restricting it to a laboratory
environment and making it difficult to track motions
within a mock-up. This drawback was at least partially
alleviated in a 2011 study through the use of a
somewhat transparent mock-up of the Orion vehicle21,22.
The mock-up allowed the cameras to see “through” the
vehicle, and motions could be tracked as subjects
performed all the mission tasks, such as
ingressing/egressing the vehicle, attaching the harnesses
and umbilical, and other tasks. The mock-up was an
innovative solution to the problems associated with
motion capture using reflective markers and cameras,
however it demonstrated the need for the ability to
capture mobility data in non-laboratory environments,
as it would have been ideal to use a higher fidelity
mock-up of Orion.
Recently, a new method of implementing IMUs has
become feasible, which has the potential to enable
mobility characterization during functional tasks in all
environments, without the need for line of sight from
expensive camera systems. Two initial studies have
recently been performed23,24 demonstrating the potential
for this methodology, which involves placing small
inertial measurement units (IMUs) onto the subject.
These trials have shown that data from the IMUs can be
converted into joint angle measurements as a subject
performs various tasks in various environments. This
methodology enables mobility measurement outside the
laboratory environment, captures motion data in three
dimensions during functional tasks, and eliminates the
need for ad hoc low fidelity vehicle mock-ups.
Inertial Measurement Units (IMUs)
In order to understand the motion of the human body
within a relevant environment (spacecraft habitable area
or spacesuit), IMUs are selected to demonstrate a novel
way of collecting data. IMUs use accelerometers and
gyroscopes to estimate relative translation and rotation.
Desirable IMU characteristics include:
• Sized to fit application (minimal mass or
• Low power consumption / long battery life;
• Dynamic range, resolution, bandwidth,
Sampling Rate, Noise, Sensitivity;
• Connection to other recording infrastructure
versus data logging / standalone;
• Comfort and/or unobtrusiveness;
• Long-term monitoring; and
• Affordable price.
This research effort has evolved from a lineage of
projects at MIT’s MVL. In order to compare the use of
IMUs for estimation of lower limb joint angles against
the standard motion capture methodology and inverse
kinematics software, a study was conducted using
commercial IMUs to capture three-dimensional
acceleration and angular velocity data generated during
human walking. Preliminary results using an extended
Kalman filter to estimate both knee and ankle joint
angles were encouraging3. Collaborators at MIT and the
Instituto Superior Técnico (Portugal) demonstrated the
efficacy of IMUs as sensory systems for gait analysis
replacing the standard motion capture camera method.
Through the use of different processing tools and
custom filtering, it was possible to improve the data
provided by the IMUs to be used for prosthetic and
orthotic devices to estimate joint kinematics during
walking25,26. On-going research implements IMU joint
kinematics in real-time for the design of ankle-foot
For an array of experimental medical and space
applications, the authors have selected IMUs that
include a set of three magnetometers, gyroscopes, and
accelerometers each. These IMUs (Opals™, APDM,
Portland, OR) are low mass wristwatch-sized devices
enabled by real-time wireless data capture or storage for
later download (see Fig. 1).
To assess IMU capabilities for human spaceflight
applications, a pilot study was conducted in a car on a
relatively smooth highway looking at constant velocity
motion and acceleration profiles. The aims of the study
were to examine ideal IMU positions and operational
protocol for data collection using a car and seat
interface as an analogue to suborbital spaceflight
keystone events simulating a seated launch23.
IAC-12-A1.6.6 Page 4 of 9
Preliminary results indicated that IMUs can be used to
characterize the human body’s motion in an analogue
situation and the vehicle’s vibrational environment.
Fig. 1: APDM IMUs are small wearable devices. Axes
are shown for IMU body reference frame.
III. PRELIMINARY VALIDATION OF IMUS VS.
MOTION CAPTURE GOLD STANDARD
To validate this data collection method we tested the
accuracy of the APDM IMUs during normal walking.
The IMUs were compared to the “gold standard” of
kinematic data collection, the Vicon motion capture
system in the Wyss Institute’s motion capture
laboratory. This system uses an array of eight T-series
cameras to track reflective markers illuminated by
infrared light. IMUs were strapped to the subject’s legs
and a plaque labelled with reflective markers was
attached (Fig. 2).
Fig. 2: IMU placement and reflective marker locations.
The locations of the reflective markers aligned with
the Y and X axis of the IMU and enabled the 8-camera
Vicon motion capture system in interpret the rotation of
the IMUs strapped to the lateral side of the upper and
lower leg. The subject was instructed to walk the length
of the motion capture volume. Limb segment rotation in
the sagittal plane was recorded with the Vicon and
APDM IMU system. Knee rotation was determined by
subtracting the rotation of the lower leg from the
reference rotation of the upper leg. A representative trial
is shown in Fig. 3.
Fig. 3: IMU-Vicon Knee walking comparison.
Areas in Fig. 3 where the Vicon data disappears
from the plot is where the subject stepped outside the
collection volume. Data was collected for 13 trials. The
average RMS error (relative to the Vicon system)
throughout the samples was 5.4 degrees with a standard
deviation of 2.4 degrees. This analysis shows that the
IMUs can be used as a substitute for a Vicon motion
capture system when an optical system is unavailable. It
also demonstrates potential advantages of the IMU
approach: freedom of movement without the restriction
of a specific motion capture volume.
El-Gohary et al. (2011)29 investigated the use of
APDM IMUs for estimating joint angles of a multi-
segment limb using a custom unscented Kalman filter
algorithms and compared data to an optical tracking
system (Eagle Analog System, Norwood, MA). All
elbow and shoulder motions analysed were found to
have IMU data correlate with greater than 0.9 to the
motion tracking system, and all cases were statistically
significant for both normal (rate not specified) and fast
(as fast as user could bend elbow) speed motions.
Another study found the APDM IMU system to
have a high Pearson's R correlation while compared to a
Vicon system (R > 0.90) for gait cadence, head rate of
rotation, and torso rate of rotation. These measurements
are typically used to test patients with mild traumatic
IV. CHAPS MEASUREMENT METHODOLOGY
Testing at David Clark Company
The APDM IMU system was brought to the David
Clark Company (Worchester, MA) to test basic mobility
in the Contingency Hypobaric Astronaut Protective Suit
(CHAPS). Before testing, it was decided to focus on the
elbow joint motion. The motion of the entire arm could
potentially be used to generate a point cloud of tracking
data to generate a work envelope, which is further
explored in the recommendations in section VII.
IMUs were placed on the forearm and bicep both
directly on the subject’s body and on the external
0 5 10 15 20 25
Knee Angle (Degrees)
Knee Flexion IMU and Vicon Comparison (RMS Error = 2.188 Degrees)
Vicon Angle Data
IMU Angle Data
IAC-12-A1.6.6 Page 5 of 9
surface of the CHAPS (two external IMUs are indicated
on Fig. 4). Additional IMUs were placed on the
fingertips (outside of glove) and on a fixed position on
the wall (not shown in figure) for future work on dead
reckoning. Three sets of motion were recorded
including: elbow flexion / extension; shoulder flexion /
extension; and shoulder abduction / adduction. The
IMUs were used to log data of the motion in two suit
pressure scenarios: unpressurized and pressurized to 1
psig. Two different methods were examined for
conducting the motion. The first method had the subject
tap the fixed wall IMU before each arm motion (tap),
and the second method had the subject move in a
continuous motion (continuous). For every trial, three
complete arm motions were conducted and the trials
were repeated twice for all sixteen scenarios (the two
methods were only used for the elbow motion).
The elbow starting position of a straight arm
(locked) of 0˚ was used in every elbow flexion /
extension trial (see Fig. 4 “zero angle reference”).
Flexion, or elbow bend, was considered positive
rotation according to the Standardization and
Terminology Committee of the International Society of
A similar study at the University of Maryland Space
Systems Laboratory24, investigated outfitting IMUs
internal to a spacesuit, using the CHAPS as a
demonstration of the technology. The Body Pose
Measurement System uses 18 IMUs on a conformal
garment worn under the suit to track body motion by
measuring the attitude of the major long bones.
Fig. 4: IMU Placement on CHAPS with elbow flexion
measurement (used with Permission from David
Results: Euler Angle Calculations
The following is an overview of the code developed
in MATLAB (The Mathworks, Natick, MA) to reduce
the acquired IMU data and find the final three Euler
angles for a pair of IMUs about a given body joint. The
basic approach is to calculate a rotation matrix that
transforms one IMU frame into another IMU frame, and
then determine the Euler angles for that rotation matrix,
which represent the 3dof rotations of the joint between
the two IMU frames. In more detail, the approach is:
1. Import data from IMU csv file to Matlab
2. Convert Quaternions to Euler Angles using
Matlab’s “quat2angle” (Aerospace tool box)
• Angles are in body reference frame X, Y,
Z with respect to North, West, Up (NWU)
world frame that the IMUs use.
3. Generate rotation matrices for both IMUs in joint
angle couple, from Bong Wie Equation 5.13 on
s1s2c3−c1s33 s1s2s3+c1c33 s1c2
c1s2c3+s1s33 c1s2s3−s1c33 c1c2
RB/A is the rotation matrix to B (NWU frame)
from A (body frame of individual IMU – A1
and A2 are used in this paper to illustrate the
two matrices for a joint angle couple as
Ri are rotation matrices about Euler angles θ1,
θ2, and θ3 that are not shown in this summary.
ci = cosθi
si = sinθi
4. Rotate the first IMU to NWU frame and then to
second IMU body reference frame using two
rotation matrices. This is done by the following
matrix chain-rule multiplication of the transpose
(inverse) of IMU-A1:
A1= rotation matrix from A1 to NWU
A2= rotation matrix from A2 to NWU
5. Compute the final three Euler angles from the
double rotation (RA1/A2) using methodology such as
G.G. Slabaugh’s white paper33. The pseudo code to
Φ ≈ 140˚
Zero Angle Reference
IAC-12-A1.6.6 Page 6 of 9
find both possible solutions* for each angle is as
θ1, ψ1, Φ1 and θ2, ψ2, Φ2 are the two Euler angle
solutions for the double rotation
Ri,j is the element in the ith row and jth column
of the 9x9 RA1/A2 matrix.
6. Data may jump from π to –π in the solution space
so “unwrap” is recommended in Matlab.
7. Zero the starting point if necessary for given joint
8. Plot rotation about primary axis (in the case of the
elbow, this was the Z axis of the IMU body frame
or Φ from the double rotation to A2 reference
V. RESULTS: CHAPS ELBOW JOINT ANGLE
The data from the four IMUs measuring the elbow
flexion inside the suit on the body and on the outside of
the CHAPS were calculated and the final z-axis Euler
angle was analysed for trends. A typical output plot is
shown in Fig. 5 that shows the internal angles of the
elbow flexion with larger values than the CHAPS. This
sample plot is from the elbow in continuous motion
with the CHAPS pressurized to 1 psig. Fig. 6 is a cross-
sectional rendering of the CHAPS and human subject,
which was developed to visually demonstrate the
*There are two solutions because of properties where
sin(π - θ) = sin(θ) and cos(θ) ≠ 0. Slabaugh explains
how to handle these in his paper.
Fig. 5: Elbow flexion data showing internal angle of
subject’s motion larger than motion of the CHAPS.
Fig. 6: Rendering of the CHAPS and human user
showing approximate differences in elbow joint
Assuming the arm started in a perfectly straight
position before every elbow flexion, the peak-to-valley
difference was measured for both internal (on the
subject’s body) and external (on the CHAPS) angles.
These maximum movement values were subtracted to
find a final difference value. Video data was also used
to compare the CHAPS external data and was found to
be similar within a few degrees (see Fig. 1 for snapshots
of elbow straight and fully bent under 1 psig conditions
from video). Video was not shot of all trials so it was
not statistically analysed. The CHAPS maximum angle
lags the internal body, but these values were not
investigated in this study.
Table 1 summarizes the statistical tests that
compared the different methodologies and suit pressure
results. If a scenario has a P value of less than 5% it is
considered significant (** denotes value is significant),
0 2 4 6 8 10 12 14 16 18 20
Elbow Angle (Degrees)
Trial 20 Elbow Flexion at 1 psig
IAC-12-A1.6.6 Page 7 of 9
meaning that the data closely match. For example, it
was found from the measurement of the internal angle,
that the two methodologies of tap versus continuous
yielded similar results (P=0.005) and that the
measurements were similar regardless of the CHAPS
pressure (P=0.002). This was not true for the external
angle, as the results of the methodology differed enough
to be not significant (P=0.128). However, the final
values of angle differences were found to be similar
regardless of method (P=0.000), but dependent on suit
Table 1: Statistical P-Values calculated to investigate
spacesuit pressure and test methodology significance
To distinguish some of the calculated averages of all
of these cases, values are presented in Table 2. Since the
method for the final angle data is significant (not much
difference between two methods) we can look at final
results and standard deviations by averaging all of the
unpressurized (26.2 ± 6.7˚) versus pressurized values
(24.2 ± 7.1˚). We see that the pressurization of the suit
leads to slightly reduced angle values of the overall
elbow flexion. The final results also show a different
story, that the continuous motion led to bigger angle
differences regardless of pressure.
Table 2: Peak angles calculated to investigate spacesuit
pressure and test methodology significance
152.1 ± 5.5
130.1 ± 4.2
22.0 ± 3.6
158.0 ± 3.5
138.4 ± 1.3
19.5 ± 3.5
157.3 ± 4.5
126.8 ± 7.3
30.5 ± 6.4
163.8 ± 3.6
134.9 ± 6.8
28.8 ± 6.9
Some key observations from this data are that the
internal angle of the human body is always larger than
the CHAPS angle (total average of all tests was 25.2 ±
6.8˚) as seen in Fig. 6; larger internal and external
angles were observed in the pressurization data versus
unpressurized; and the continuous method had larger
angles for the internal angle and smaller for the external.
VI. LIMITATIONS OF IMUS
IMU systems have limitations and the optimal
system must be selected for the right job. Typical
limitations are in g-range, sensitivity, lag, filtering, and
accuracy. A few issues were identified with the selected
APDM system for this application and are described in
The investigators found that the magnetometers
were susceptible to magnetic interference, even from
metal tabletops, which changes the orientation of the
NWU coordinate frame. For the CHAPS testing there
was little magnetic interference, but this should be
monitored in all testing environments and can be
displayed with custom Matlab code to show the data in
real time. A study by Bachmann et al. in 2004
developed a guideline that errors can be avoided by
maintaining an approximate distance of two feet from
any source of disturbance29.
For the elbow joint, the positioning of the IMUs was
closely matched internally and externally, but as seen in
Fig. 1, the IMUs are not exactly on the rotation axes of
the arm. For this reason, it was desired to find the final
three Euler angles before reducing any data. The final
rotation axis data may therefore have some twist
associated with the values and this data should be used
as a proof of concept.
The CHAPS was not sized specifically for the
subject in this test, and the suit is designed to be used
nominally in the seated position. Had the experiments
been performed with a perfectly fitting suit in the seated
position, it is possible that the internal and external
measurements would be more similar.
Dead reckoning is a technological issue for IMUs as
they do not know their exact positions at any given
time. This is why the “tap” methodology was tested, to
try and have a reset point in physical space. The
advancement of this technique is further explored in the
next section. A common indicator of the difficultly of
position tracking is from the occurrence of drift.
Pseudo markers can be estimated from video
analysis. However, in future testing it would be ideal to
have arm markers for validation in photos or videos.
VII. RECOMMENDATIONS FOR FUTURE
The following are recommendations for future
improved data acquisition testing with IMU systems for
applications like spacesuit mobility.
IMU data could be verified by constructing a simple
non-ferrous rig to test one degree of freedom at a time
with known angles and potentially known rotation rates
IAC-12-A1.6.6 Page 8 of 9
Real time acquisition of Euler angles can be
generated with Matlab code. The signal will be slightly
lagged, but the instant validation of motion would be
useful and more insightful. This could also be used for
real time monitoring of a variety of spacesuit joints.
Position estimation would be a valuable addition to
the capabilities of the IMU. The performance of the
ADPM IMUs is marginal for position estimation, due to
drift, without regular position fixes. ADPM has
unreleased code that uses frequent position fixes (every
5 seconds) and velocity nulling to track IMU position;
performance figures have not been released but a
comparison of the estimate to video of an IMU is
compelling. We attempted some trials using arm motion
in which position fixes were provided using a tap
between two IMUs. These events can be identified and
used as position fixes. Analysis of these trials is on-
going. If adequate performance can be demonstrated,
either through position fixes and careful software
correction, or via future hardware improvements, IMU
position estimation would enable a variety of
applications such as:
• Generating a point cloud that maps out space suit
workspace envelopes using natural motions.
• Tracking displacement in all directions for
standardized tasks such as using a tool, useful for
tool and task optimization.
• Enabling more general motion capture without the
cost and constraints of a vision-based system.
It is also recommended that future work in spacesuit
motion tracking incorporate El-Gohary et al.’s (2011)
This proof of concept research met the goal of
demonstrating that measurements of the human body
within a spacesuit can be taken in a novel method using
inertial measurement units (IMUs). With IMUs it is
possible to track internal versus external angles to figure
out optimal spacesuit fit, energy expenditure, and work
envelope. Refinement of the method should prove to be
valuable while testing in analogue environments or out
in the field without the need for a visual motion capture
system. Future data could be collected during
spaceflight and lead to improved spacesuit design.
The authors would like to acknowledge support from
the Man-Vehicle Laboratory and Dr. Alan Natapoff, the
David Clark Company, especially Donald B. Tufts, the
Wyss Institute for Biologically Inspired Engineering at
Harvard, Gabe Montague for CHAPS renderings and
animations, and Massimiliano Di Capua for reviewing
this manuscript. This research was made possible in part
by the MIT Portugal Program.
1. Favre, J., Aissaoui, R., Jolles, B.M., Guise, J.A.,
and Aminian, K. (2009): “Functional calibration
procedure for 3D knee joint angle description using
inertial sensors”. Journal of Biomechanics. Vol. 42,
2. Schepers, H.M., Koopman, H.F.J.M., and Veltink,
P.H. (2007): “Ambulatory Assessment of Ankle
and Foot Dynamics”. IEEE Transactions on
Biomedical Engineering. Vol. 54, Pp. 8995.
3. Young, D., D’Orey, M.S.R., Opperman, R.,
Hainley, C., and Newman, D.J. (2010): “Estimation
of Lower Limb Joint Angles During Walking Using
Extended Kalman Filtering”. 6th World Congress of
Biomechanics. IFMBE Proceedings. Vol. 31, Pp.
4. Young, D., and Newman, D.J. (2011): “Chapter:
Augmenting Exploration: Aerospace, Sea and
Self. Wearable Monitoring Systems”. Eds. Annalisa
Bonfiglio and Danilo De Rossi. Springer Science.
5. Moreno, J.C., Lima, E.R., Ruíz, A.F., Brunetti, F.J.,
and Pons, J.L. (2006): “Design and implementation
of an inertial measurement unit for control of
artificial limbs: Application on leg orthoses.”
Sensors and Actuators B: Chemical. Vol. 118, Pp.
6. Tee, K.S., Awad, M., Dehghani, A., Moser, D., and
Zahedi, S. (2011): “A Portable Gait Monitoring
System for Lower Limb Prosthetic Alignment”.
World Congress on Engineering. Vol. 3, Pp. 2131-
7. Teskey, W.J.E., Elhabiby, M., and El-Sheimy, N.
(2012): “Inertial Sensing to Determine Movement
Disorder Motion Present before and after
Treatment”. Sensors. Vol. 12, Pp. 3512-3527.
8. Callaway, A.J., Cobb, J.E., and Jones, I. (2009): “A
Comparison of Video and Accelerometer Based
Approaches Applied to Performance Monitoring in
Swimming”. International Journal of Sports
Science & Coaching. Vol. 4, Pp. 139-153.
9. Lapinski, M. (2009): “A Distributed Wearable,
Wireless Sensor System for Evaluating Professional
Baseball Pitchers and Batters”. 13th IEEE
International Symposium on Wearable Computers.
10. Klaus, D.M., Metts, J., Kobrick, R., Mesloh, M.,
Monk, T., Gauthier, E., Eberhart, K., Baca, D.,
Wright, C., Gustafson, A., Oryshchyn, L., and
Massey, D (2007): “Space Suit Concepts and
Vehicle Interfaces for the Constellation Program”.
SAE Technical Paper 2007-01-3088.
11. Lee, G.R., Graziosi, D., Splawn, K., and Ferl, J.
(2011): “Enhancing Capability in Launch, Entry,
and Abort Style Spacesuits Fro ISS and
Commercial Use”. 41st International Conference on
Environmental Systems. AIAA 2011-5032.
IAC-12-A1.6.6 Page 9 of 9
12. Jacobs, S.E, and Tufts, D.B. (2011): “Follow-On
Development of the Demonstrator Suit for Post-
Shuttle Operations”. 41st International Conference
on Environmental Systems. AIAA 2011-5030.
13. Newman, D.J., Canina, M., and Trotti, G.L. (2007):
“Revolutionary Design for Astronaut Exploration –
Beyond the Bio-Suit System”. 4th Symposium on
New Frontiers and Future Concepts, Space
Technology and Applications International Forum.
American Institute of Physics. Pp. 975-986.
14. Aitchison, L.T. (2012): “A Comparison of Methods
for Assessing Space Suit Joint Ranges of Motion”.
42nd International Conference on Environmental
Systems. AIAA 2012-3534.
15. Jones, R.L. (1966): “Evaluation and Comparison of
Three Space Suit Assemblies.” NASA Technical
16. Ross, A. (2000): “Advanced Space Suit Isolated
Joint Mobility Test for the Space Suit Comparative
Technology Evaluation Test” JSC-39522 (CTSD-
17. Abramov, I., Moiseyev, N., and Stoklitshy, A.
(2004): “Some Problems of Selection and
Evaluation of the Martian Suit Enclosure Concept”.
Acta Astronautica, 57, 901-909.
18. Tufts, D.B., Jacobs, S.E., and Barry, D.M. (2010):
“The Demonstrator Suit: Incorporating Waist, Hip
and Shoulder Mobility in a Full Pressure Suit” 40th
International Conference on Environmental
Systems. AIAA 2010-6178.
19. Ripps, T.B, Garcia, J.M., Macleod, S.M., and
Jacobs, S.E. (2011): “The Demonstrator Suit:
Evaluating a Full Pressure Suit through Manned
and Unmanned Testing”. 41st International
Conference on Environmental Systems. AIAA
20. England, S., Benson, E., and Rajulu, S. (2010):
“Functional Mobility Testing: Quantification of
Functionally Utilized Mobility among Unsuited and
Suited Subjects”. Report No. NASA/TP-2010-
21. Jacobs, S.E., and Tufts, D.B. (2012): “Space Suit
Development for the Multi-Purpose Crew Vehicle”.
42nd International Conference on Environmental
Systems. AIAA 2012-3550.
22. Ripps, T.B, Garcia, J.M., and Macleod, S.M.
(2012): “Evaluating Suited Ingress and Egress of a
Space Vehicle (Orion) Seat”. 42nd International
Conference on Environmental Systems. AIAA-
23. Kobrick, R.L., Carr, C.E., Meyen, F., Domingues,
A.R., and Newman, D.J. (2012): "Using inertial
measurement units for space vehicle safety,
comfort, design, and performance optimization"
(abstract and presentation). Next-Generation
Suborbital Researchers Conference 2012. Session:
Life Sciences. Palo Alto, California, USA. Paper
Registration ID: 38594404.
24. Di Capua, M., and Akin, D. (2012): “Body Pose
Measurement System (BPMS): An Inertial Motion
Capture System for Biomechanics Analysis and
Robot Control from Within a Pressure Suit”. 42nd
International Conference on Environmental
25. D’Orey, M.S.R., Martins, J.M., Silva, M.T., and
Newman, D.J. (2011): “Detection of Gait Cycle
Events Through Fourier Series Expansion of
Kinematic Data”. EUROMECH Colloquium 511
on Biomechanics of Human Motion.
26. D’Orey, M.S.R. (2012): “Passive Dynamic Walkers
and Sensory Systems for Gait Analysis”. Ph.D.
Thesis, Instituto Superior Técnico and
Massachusetts Institute of Technology (MIT
27. Domingues, A.R., Marreiros, S.P., Martins, J.M.,
Silva, M.T., and Newman, D.J. (2011): “Skin Strain
Field Analysis of the Human Ankle Joint”, 4º
Congresso Nacional de Biomecânica (CNB2011).
28. Domingues, A.R., Marreiros, S.P., Martins, J.M.,
Silva, M.T., and Newman, D.J. (2012): “Analysis
of Ankle Skin Deformation for the Development of
Soft Orthotics” (abstract). 18th Cingress of the
European Society of Biomechanics (ESB2012).
Paper ID: 01713.
29. El-Gohary, M., Holmstrom, L., Huisinga, J., King,
E., McNames, J., and Horak, F. (2011): “Upper
limb joint angle tracking with inertial sensors”.
Conference Proceedings, 33rd Annual International
Conference of the IEEE Engineering in Medicine
and Biology Society. Pp. 5629-5632.
30. Simoes, M.A. (2011): “Feasibility of Wearable
Sensors to Determine Gait Parameters”. Master’s
Thesis. University of South Florida, Graduate
School Theses and Dissertations. Paper 3346.
31. Wu, G., van der Helm, F.C.T., Veeger, H.E.J.,
Makhsous, M., Van Roy, P., Angling, C., Nagels,
J., Karduna, A.R., McQuade, K., Wang, X.,
Werner, F.W., and Buchholz, B. (2005): “ISB
recommendation on definitions of joint coordinate
systems of various joints for the reporting of human
joint motion—Part II: shoulder, elbow, wrist and
hand”. Journal of Biomechanics, 38, Pp. 981-992.
32. Wie, B. (1998): “Space Vehicle Dynamics and
Control”. Page 311, Equation 5.13. AIAA
Educational Series, Reston, VA.
33. Slabaugh, G.G. (1999): “Computing Euler angles
from a rotation matrix”. Report from personal
website. Available online (last accessed 5 July