Content uploaded by Falin Wu
Author content
All content in this area was uploaded by Falin Wu on Apr 07, 2020
Content may be subject to copyright.
CHALLENGING OLYMPIC MEDALS – AN INNOVATIVE GNSS-BASED MULTI-
SENSOR SYSTEM FOR ATHLETE TRAINING AND COACHING
K. Zhanga, F. Wua, C. Mackintoshb, T. Riceb, C. Goreb, A. Hahnb, and S. Holthousc
aSchool of Mathematical and Geospatial Sciences, RMIT University, Melbourne, Australia
bAustralian Institute of Sport, Canberra, Australia
cCatapult Innovations Pty Ltd, Melbourne, Australia
KEY WORDS: Global Positioning System, Olympic Competition, Inertial Navigation System, Sport Application
ABSTRACT:
In preparing for Olympic competition, every detail is important. Athletes and coaches are not only interested in the physiological
information, but also interested in the positional information. Physiological information can be relatively easy to obtain using
relevant detectors. However, real-time high precision determination of the athletes' movement and the status of motion has been a
challenging task. This contribution introduces the development of a smart real-time athlete monitoring and coaching system, which
integrates a low-cost GPS, Micro-Electro-Mechanical System (MEMS) inertial measurement units (IMUs), magnetometers, wireless
communication and other physiological sensors. Surveying, Positioning And Navigation (SPAN) group at RMIT University in
collaboration with the Australian Institute of Sport and Catapult Innovations Pty Ltd. has embarked this innovative development of
the patented system since early 2003. The system was initially designed for elite Australian rowers in both Athens Olympics and the
Lucerne International Rowing Regatta and now it has been tested and used across a large number of sport activities, including
rowing, canoeing, skiing, running, sailing, footballing etc. The current developments of the system, system architecture, field
validation and potential applications are presented and further developments (e.g. the interpretation of data and customisation for
individual players) and its commercial aspect of the project are also outlined. We hope that the 2008 Beijing Olympic Games will
witness the significant roles of geospatial science has played in sport competition and quality improvement of recreational activities.
1. INTRODUCTION
In many sports, the margin between victory and defeat may be a
matter of a few hundredths of a second. Certainly that is true of
sport competition where the demands on equipment and the
performance pressure on athletes are tremendous - and not just
on elite athletes, but increasingly on participants at every skill
level. Athletes and coaches are not only interested in the
physiological information, e.g., blood oxygen, respiration and
heart rates; but also interested in the position and movement
information, e.g. position, velocity, acceleration and changes in
direction. Physiological information can be relatively easy to
obtain using relevant detectors, e.g., heart rate monitor,
ergometer (O'Sullivan et al., 2003). However, real-time high
precision determination of the athlete location and movement
has been a challenge (Fyfe et al., 2001; Hutchings et al., 2000;
Larsson, 2003; Wu et al., 2007).
In-depth understanding of sensor-based human performance
measurement, such as the determination of a characteristic
signature of the "perfect" movement, is required to analyse the
performance of the athlete (Seiler, 2003). The position,
movement (velocity and heading) and acceleration (i.e. force)
information plays an important role in an effective analysis of
the athlete performance. Athletes and coaches are not only
interested in the trajectories of the movement, but also in the
motion analysis of segments of the human or the orientation of
equipment (e.g. lift-over of a motor cycle, torsion of skis).
Therefore, the ability to measure and record positional
information together with athlete physiological information in
real-time is critical to the process of athlete training and
coaching (Wu et al., 2007).
Traditionally, development and testing of materials or
equipment for sport has been based on repeated measurements
with resources including timing cells or wind tunnels. Similarly,
the analysis of athletes’ performance has relied on techniques
such as measuring race segments (chronometry) and video
recordings (Wägli and Skaloud, 2007). That is to say, positional
information can only be measured in either well-controlled
situation in dedicated sport laboratories or using simulation
device (Zhang et al., 2004). These methods, however, appear
vulnerable to changing meteorological conditions and the
difficulty of replicating the posture and movement of test
subjects from one trial to the next due to such facts as improved
performance stemming from cumulative experience in the trials
or decreased performance due to fatigue (Wägli and Skaloud,
2007). Furthermore, much of the equipment is either too heavy,
expensive or obtrusive and multiple factors which are difficult
to control have limited the use of sport-specific field testing
(Zhang et al., 2003). Therefore, new methods that offer precise
measurements during trials and subsequent evaluation of
positions, velocities, acceleration and changes in direction
would be a big leap forward (Wu et al., 2007).
Satellite-based positioning has already proven its effectiveness
in many sports, including car racing and rowing (Edgecomb and
Norton, 2006; Wägli and Skaloud, 2007; Zhang et al., 2003;
Zhang et al., 2004). In an open space, Global Positioning
System (GPS) supports continuous position, velocity, and
acceleration analysis of athlete’s trajectories. The
environmental reality, however, is often far from such an ideal
case. An athlete’s environment quickly alternates between open
spaces and areas that block or attenuate satellite signals (e.g.
sudden satellite masking), which makes GPS signal reception
difficult or even impossible. To overcome the lack of continuity
of the GPS signals and in order to observe accelerations (and
hence forces) directly, low-cost MEMS-IMUs are integrated
1799
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
with GPS. Such a combination is suitable for sport’s application
because of their small size and limited cost. Also, the
GPS/MEMS-IMU integration enables accurate determination of
the position, velocity and acceleration (Grewal et al., 2007;
Titterton and Weston, 2004).
Surveying, Positioning And Navigation (SPAN) group at RMIT
University in collaboration with the Australian Institute of Sport
and Catapult Innovation Pty Ltd. has embarked an innovative
development of a smart real-time athlete monitoring and
coaching system since early 2003. This paper presents recent
developments of the integrated sensors system. Performance of
the low-cost code-only and carrier phase GPS modules is
evaluated using high-end GPS systems. Two prototype systems
are then introduced. Finally, a typical sport application is
illustrated.
2. DEVELOPMENT OF PROTOTYPE SYSTEM
This research commenced in 2001 from the Australian
Cooperative Research Centre (CRC) for Micro Technology,
under Project 2.5 "Interface Technologies for athlete
Monitoring". Established in July 1999 with a seven year grant,
the CRC for Micro Technology has four major research areas:
fabrication technology; microdevice packages; safety and health;
and micro-fluidic devices (CRC microTechnology, 2003). The
aim of the Project 2.5 is to develop unique monitoring
equipment that is essentially unobtrusive, so that the athlete is
virtually unaware of its presence in training and competition.
The Project 2.5 initially investigates the feasibility using GPS
to aid inertial devices for position, velocity and acceleration
(PVA) determination in real-time. The positional information is
then combined with other athlete physiological information and
integrated into a dedicated electronic device for package and
analysis, and relayed to the coach. The continuous monitoring
of three-dimensional PVA of the athlete with a very high update
rate and accuracy is achieved through an integration of GPS, an
IMU/INS for athlete physiological information, a data
communication mechanism, and interactive visual control using
a Geographical Information System (GIS) (Wu et al., 2007).
Table 1 outlines the major phases of the system development
(Wu et al., 2007). The Stages I, II, III and IV have been
completed and stages V is under intensive development. In
stages I, II and III, the performance of low-cost GPS has been
evaluated using high-end GPS receivers. In stage IV, a
prototype system has been developed and used in rowing
training and coaching. In stage V, an upgraded system is
developed. The GPS chip, MEMS IMU and wireless
communication device have been integrated in the final
(compact) system. Online calculation and GIS services will be
also integrated in the future development to make the system
user friendly.
2.1 The Sport of Rowing
Rowing is a highly developed and becomes an increasingly
popular international sport. It combines a wonderful spectacle
with a heated competition. Rowing races usually cover a
distance of 2,000 m in river, canal or lake-type competition
environment in six lanes. To win the competition, athletes have
to qualify through four pre-determined rounds: the preliminary
round (heats), the repeat round (repechages), the semi-finals
and the finals. The "A" final determines the first six places and
the runners-up "B" final determines the next six places (i.e. 7th
to 12th positions). The number of rounds per event depends on
the number of crews taking part.
Stage Hardware Software PVAT
I iPAQ+PC, Genius
1, support circuit,
active/patch antenna
WCE and
Win2000/
XP based
software
Post-
processing,
low update
rate
II iPAQ, Genius 1,
support circuit,
active/patch antenna
WCE
based
software
Real-time,
low update
rate
III MCU, SuperStarII,
MCU built-in
support circuit,
patch antenna
MCU
based
firmware
Real-time,
high update
rate
IV
ROVER-2004:
Chip-level, Fastrax
GPS chips, built-in
circuit, active/patch
antenna, wireless
communication
MCU
based
firmware
Real-time,
high update
rate
V minimaxX:
Chip-level, GPS,
IMU and
magnetometer
chips, built-in
circuit, active/patch
antenna, wireless
communication
Win XP
based
software /
MCU
based
firmware
Real-time,
high update
rate
Table 1. Major development milestones of the smart integrated
tracking system (PVAT – Position, velocity,
acceleration and time)
The races are judged under the supervision of umpires, who are
members of the Jury for the event. The jury members are placed
at various locations on and off the competition course, such as
the starting line, where the races begin under the supervision of
the aligner and the starter; along the course of the race in the
competition lanes under the supervision of umpires; the
finishing line with the finish-line umpire; the identity
verification stage of the crews before their embarkation onto the
boats; the weighing-in of the athletes; the weighing-in of boats;
and, in general, in all areas directly related to the competition,
the athletes and their equipment (Athens Olympics, 2004).
(a) Sculling events (b) Sweep-oared events
Figure 1. Sculling (two oars used, one in each hand) and sweep-
oared (one oar with both hands) scenarios in Olympic
competition
There are 14 different boat classes raced in Olympic rowing.
These include eight sculling events in which two oars are used
(see Figure 1a), one in each hand and six sweep-oared events in
which the rowers use one oar with both hands (see Figure 1b).
1800
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
The sculling boat classes are the single, the double and the
quadruple sculls with crews of one, two or four athletes
respectively, as well as the lightweight double. The sweep row
categories include the pair, the four, the lightweight four (for
men only) and the eight with coxswain.
2.2 Rowing Phases and Monitoring
A rowing stroke is a precise movement with rowers using their
legs, back and arms to generate power. A stroke begins with the
placing of the blade in the water and ends with the re-emerging
of the blade from the water and positioning for another cycle.
The rowing stroke can be divided into four main phases: catch,
drive, finish and recovery (Isaacs and Mulligan, 1999; Rower's
World, 2003) (see Figure 2). These sequential phases must flow
from and into each other to produce a continuous and fluid
movement.
Figure 2. Schematic diagrams showing the four main sequential
phases (i.e. catch, drive, finish and recovery) in a
rowing stroke and the position of head, arm and legs of
the rower).
At the catch, the blade is placed into the water quickly with
minimal disturbance to the boat. The rower's arms are extended
outward, torso is tilted forward, and legs are compressed. A
good catch produces a minimal amount of back and front splash
and causes no check. The catches of all crews of a boat must be
identical. Out of step catches (unsynchronisation) causes
balance problems and reduce a boat's speed. The blade must be
fully squared to the water at the catch (ibid).
The boat gains its speed on the drive. In this part of the stroke,
the oarsman applies power to the oar with forces from arms,
back and legs, and swings his torso away from the stern of the
boat. The handle of the oar is pulled in a clean, powerful and
levelled motion towards the bow of the boat with a constant
force.
At the finish, the oarsman finishes applying power to the oar
handle, removes the blade from the water sharply, and feathers
the oar (rotate it by 90º) so that the blade becomes parallel to
the surface of the water.
At the recovery, rowers are given a brief rest to prepare for the
next stroke. The oarsman must slide towards the stern of the
boat and prepare the blade for the next catch. Crews exhibit an
approximate 2:1 ratio between the times spent on the recovery
and the times spent on the drive. At the end of the recovery, the
oar is gradually squared and prepared for the catch (Mickelson,
1979).
Understanding which movements should occur in each phase of
the stroke allows coaches to design effective conditioning
programs and evaluate rowing performance effectively (Seiler,
2003). Success in competitive rowing is achieved by taking the
shortest time to complete a course (usually 2000 m) which
directly links to the average velocity of the boat. Acceleration
is proportional to force since the boat is accelerated as it reacts
with the sweeping arc of the oar. Three factors affecting boat
velocity, power, length and rate, are important determinants of
rowing performance. The power provides how fast the boat
travels in a stroke, the length is associated with how far the boat
travels in each stroke and the rate striking provides how many
strokes are rowed per minute (Xiao et al., 2003). Therefore the
rower must achieve an optimal combination of high stroke
power, long stroke length and high stroke rate.
Figure 3 presents the stroke signals captured using geodetic-
type GPS receivers. It is demonstrated that the signals captured
provide a clear picture of the rowing stroke phases as described
above. In this particular stroke, the graph indicates that the
rower has harmonised well in his stroke cycle by using
appropriate time (1:2) in the catch and the drive.
Figure 3. Schematic rowing stroke signature captured from
high precision GPS measurements
2.3 Evaluation of Low-cost, Low Update Rate Code GPS
A number of critical factors contribute to the applicability of the
GPS system to sport applications. This includes the precision,
cost, volume and weight of the system, and its integration with
other sensors including accelerometers, communication
mechanism, and a personal digital assistant (PDA). Because of
the challenges of diverse applications and a broadening market,
the low-cost GPS has evolved significantly during the past
decade. Increasingly miniaturised device has been developed
into even wearable and embedded into other devices. For
financial consideration the tracking of a large number of
athletes requires the use of low-cost, single-frequency GPS
receivers. The current pricing of dual-frequency GPS receivers
means that their use will be restricted to a few athletes and
applications with high position-accuracy requirements (Wägli
and Skaloud, 2007). Thus, the research presented here
integrates low-cost single frequency GPS with other sensors
that take into consideration the high dynamics of the athletes
and their particular environment.
An important component of the research is to evulate the
performance of a low-code GPS receiver before it is integrated
into a sport application system. A rowboat test has been
conducted to compare the low-cost, code-only GPS receiver
with high-end dual-frequency receiver, Trimble 5700. The low-
cost, code-only GPS receiver used is a Genius 1 manufactured
by Rojone (2003), and based on SiRF GPS chipset. These two
types of GPS receiver are mounted on a rowboat and tested in a
rowing race course in Canberra, Australia.
Figure 4 shows the differences in velocity determined
simultaneously from the low-cost code-only GPS and the high-
end dual-frequency GPS. Assuming the carrier velocity from
1801
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
Post-Processing Kinematic (PPK) derived velocity has an
average accuracy in the order of ~0.03 m/s. The results confirm
that the accuracy specification of 0.1 m/s from the manufacturer
is correct for more than 95% of observations. It is demonstrated
that low-cost code-only GPS receivers can provide accuracy
position, velocity and acceleration information.
Figure 4. GPS velocity differences between code and RTK
solution
2.4 Evaluation of Low-cost, High Update Rate Carrier
Phase GPS
An athlete monitoring and coaching system with the
functionality to output PVA information at a high update rate
has been found necessary and essential in many sport
applications. The GPS velocity and acceleration can be derived
either in the position domain or the range domain using a
differentiator. As the GPS position is usually derived from the
pseudorange measurements, the position error is of the order of
several to tens of metres. The derived velocity and acceleration
are therefore very noisy and inaccurate. The position solution
derived from carrier phase measurements is much more
accurate and less noisy. But such a solution needs multiple
dual-frequency receivers, i.e. through forming double-
differenced carrier phase measurements to derive the baseline
solution. Such an implementation is too expensive and awkward
for a MEMS-based sensor system. The implementation in the
range domain is relatively sophisticated. The velocity and
acceleration can be derived from the Doppler measurements or
from the carrier phase measurements (Li et al., 2006). Another
benefit of the low-cost GPS receiver with carrier phase
measurement is that the accuracy of position can be improved
using the carrier smoothing technique. Carrier phase smoothing
is a process that combines the absolute but noisy pseudorange
measurements with the accurate but ambiguous carrier phase
measurements to obtain a good solution without the noise
inherent in pseudorange tracking through a weighted averaging
process (Misra and Enge, 2006).
To evaluate the performance of the low-cost, high update rate
carrier phase GPS, a mini-bus experiment is conducted in Yarra
Bend Park, Melbourne in 2003. A RTK GPS system (Trimble
5700 receivers) is again used for a “ground-truth” reference.
Obviously, the RTK solution is precise enough for this purpose
when evaluating the 10 Hz update rate GPS solution. A low-
cost, high update rate, carrier phase GPS is used in this system
to obtain the range velocity and acceleration. A Canadian
Marconi Company’s (CMC) SuperStar II GPS OEM board
(Novatel Inc., 2007) is used to form the hardware basis of the
system. The SuperStar II can provide PVA solution at a rate up
to 5 Hz as well as raw measurements (i.e. code, carrier phase
and signal-to-noise ratio) at a maximum rate of 10 Hz.
Figure 5 represents the position discrepancies (latitude and
longitude) between solutions derived from code-only and
carrier-smoothed measurements respectively (compared with
RTK solution again). The carrier-smoothed solution has
smoothing lines and the code-only solution is noisy. Figure 6
depicts the difference of two velocity solutions derived from the
10 Hz GPS receiver and RTK GPS solution respecively. The
result shows that the average and stand deviation of the
difference are 0.005 m/s and 0.088 m/s respectively. Because
the obstruction of buildings and trees in the urban environment,
the discontinuity of GPS solutions is evident.
Figure 5. Position differences between 10 Hz low-cost GPS
(code-only and carrier-smoothed solutions) and RTK
GPS solutions
Figure 6. Velocity differences between 10 Hz low-cost GPS
and RTK GPS solutions
It has clearly demonstrated that carrier-smoothed solution can
reduce the level of noise of code-only solution efficiently.
However, slow-variation errors such as SVs' ephemeris and
atmosphere errors are still present in the carrier-smoothed
solution.
2.5 Prototype Athlete Monitoring and Coaching Systems –
ROVER-2004 and minimaxX
In 2004, a prototype athlete monitoring and coaching system,
namely ROVER-2004, was developed initially for rowing
training and coaching (Zhang et al., 2004). Figure 7 shows the
prototype system, ROVER-2004, which integrates a low-cost
GPS and wireless communication. The GPS receiver used is a
Fastrax iTrax03 chip which is a low-cost single frequency
receiver with carrier phase measurements.
1802
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
Robust algorithms and associated software/firmware have been
developed using a number of special treatments to suit for
rowing-specific applications. These algorithms are evaluated
through a number of static and kinematical trials using high
precision RTK-GPS (Wu et al., 2007).
Figure 7. A compact prototype rower monitoring and coaching
system (ROVER-2004) with 10 Hz carrier GPS
capability and wireless communication.
As aforementioned, an athlete’s environment can change
dramatically between open spaces and areas where GPS signal
reception is difficult or even impossible due to obstruction. To
overcome the problem of GPS signal continuity and observe
accelerations directly, low-cost MEMS IMUs are integrated
with GPS chips. A new version of athlete monitoring and
coaching system, namely minimaxX, has been developed (Wu
et al., 2007).
Figure 8. A compact prototype athlete monitoring and coaching
system (minimaxX) with high update rate carrier phase
GPS capability and wireless communication
The minimaxX is designed as a low-cost, wearable and
integrated positioning device including a low-cost GPS, MEMS
IMUs, and three magnetometers aligned with three orthogonal
axes. The dimension of the minimaxX is approximately
8¯5¯4 cm (see Figure 8). The GPS receiver is a Fastrax
iTrax03 chip (Fastrax, 2007).
3. SPORT APPLICATIONS
The ROVER-2004/minimaxX systems were initially designed
for elite Australian athletes in both Athens Olympics and the
Lucerne International Rowing Regatta and now it has been used
across a large number of sport activities, including rowing,
canoeing, skiing, running, sailing, cycling and footballing etc.
(see Figure 9) (Wu et al., 2007; Zhang et al., 2004). The latest,
but typical Australian Football League application of the
minimaxX was launched in May 2007 (Wu et al., 2007). A
stream of new AFL specific functionalities has been developed.
For example, the minimaxX system has been coupled with a
powerful on board computer and long range wireless
communication for up to 100 units simultaneously, and it can
broadcast performance summaries for coaches or the media to
use in real time. Specific functionality has been developed for
real-time monitoring and post-training analyses of footballers.
The following discussion is concentrated on rowing application
in Olympics.
Figure 9. Some typical examples of ROVER-2004/minimaxX
systems’ sport applications (e.g. sailing, canoeing,
footballing, half pipe, rowing, mogul skiing, cycling
and skiing)
Figure 10. James Tomkins and Drew Ginn who won men’s pair
gold medal in Athens Olympics
The Athens 2004 Olympic Games Rowing events were held at
the Schinias Olympic Rowing and Canoeing Centre over a
period of nine competition days, from 14 to 22 August 2004. A
total of 550 athletes (358 men and 192 women) from all over
the world took part in 14 rowing events. 45 (28 men and 17
women) Australian rowers took part in 11 rowing events
(Athens Olympics, 2004). The ROVER-2004 was used by
Australian rowers prior to and during the Athens Olympic
Games. Three Olympic rowing medals were won by Australian
athletes. Figure 10 shows James Tomkins and Drew Ginn who
won men’s pair gold medal in Athens Olympics. The Beijing
2008 Olympic Games Rowing events will be held at Shunyi
Olympic Rowing-Canoeing Park on the side of Chaobai River
in Shunyi District of Beijing from August 9-17, 2008. A total of
1803
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008
550 athletes (350 men and 194 women) from all over the world
will take part in 14 rowing events (Beijing Olympics, 2008).
We hope the 2008 Beijing Olympic Games will witness the
significant roles of geospatial science has played in sport
competition and improvement of our quality of recreational
activities
4. CONCLUSION
This paper presents the latest developments of a smart real-time
athlete monitoring and coaching system, which integrated a
low-cost GPS, MEMS IMUs, magnetometers and wireless
communication. The current developments of the system,
system design, validation and applications have been outlined.
The proof-of-concept trials are successful and the feasibility of
GPS technology to assist in elite athlete training has been
confirmed. Prototype systems have been developed and used at
the elite level across a large number of sports activities.
ACKNOWLEDGMENTS
This research is supported by the Australian Institute of Sport
Discretionary B Funding.
REFERENCES
Athens Olympics, 2004, Athens Olympic Rowing,
http://www.athens2004.com/en/RowingRules.
Beijing Olympics, 2008, Rowing - The Beijing Olympic Games,
http://en.beijing2008.cn/cptvenues/sports/rowing/index.shtml.
CRC microTechnology, 2003, Cooperative Research Centre
(CRC) for microTechnology,
http://www.microtechnologycrc.com/.
Edgecomb, S. J., and Norton, K. I., 2006, Comparison of Global
Positioning and Computer-based Tracking Systems for
Measuring Player Movement Distance during Australian
Football: Journal of Science and Medicine in Sport, v. 9, no. 1-
2, p. 25-32.
Fastrax, 2007, iTrax03 OEM GPS Receiver Module Data Sheet,
http://www.fastrax.fi/index.cfm?template=products.show.cfm&
guid=4594da1a-503c-469c-91b2-6948043189be.
Fyfe, K. R., Rooney, J. K., and Fyfe, K. W., 2001, Motion
Anaysis System: US Patent Office, US6301964.
Grewal, M. S., Weill, L. R., and Andrews, A. P., 2007, Global
Positioning Systems, Inertial Navigation, and Integration, John
Wiley & Sons, Inc.
Hutchings, L. J., Gross, R., and Jarpe, S., 2000, System and
Method for Measuring Movement of Objects: US Patent Office,
US6122960.
Isaacs, N., and Mulligan, S., 1999, Rowing 101, The University
of Washington Student Newspaper:
http://archives.thedaily.washington.edu/1999/060299/S5.rowing
.html.
Larsson, P., 2003, Global Positioning System and Sport-
Specific Testing: Sports Medicine, v. 33, no. 15, p. 1093-1101.
Li, Y., Dempster, A., Li, B., Wang, J., and Rizos, C., 2006, A
Low-cost Attitude Heading Reference System by Combination
of GPS and Magnetometers and MEMS Inertial Sensors for
Mobile Applications: Journal of Global Positioning Systems, v.
5, no. 1-2, p. 88-95.
Mickelson, T. C., 1979, Functional Anatomy of the Rowing
Stroke: Oarsman, v. 11, no. 3, p. 6-9.
Misra, P., and Enge, P., 2006, GLOBAL POSITIONING
SYSTEM, Signals, Measurements, and Performance: Lincoln,
Massachusetts, USA, Ganga-Jamuna Press.
Novatel Inc., 2007, SuperStar II,
http://www.novatel.com/products/superstar.htm.
O'Sullivan, F., O'Sullivan, J., Bull, A. M. J., and McGregor, A.
H., 2003, Modelling Multivariate Biomechanical Measurements
of the Spine during a Rowing Exercise: Clinical Biomechanics,
v. 18, no. 2003, p. 488-493.
Rojone, 2003, Rojone Pty. Ltd,
http://www.rojone.com.au/about.html.
Rower's World, 2003, The Stroke, About Rowing,
http://www.rowersworld.com/Content/About/2000_7_10n.php.
Seiler, S., 2003, Rowing Physiology and Performance,
http://home.hia.no/~stephens/rowing.htm.
Titterton, D. H., and Weston, J. L., 2004, Strapdown Inertial
Navigation Technology, IEE Radar, Sonar, Navigation and
Avionics Series: Stevenage, The Institution of Electrical
Engineers and The American Institute of Aeronautics and
Astronautics.
Wägli, A., and Skaloud, J., 2007, Turning Point - Trajectory
Analysis for Skiers, Inside GNSS, p. 24-34.
Wu, F., Zhang, K., Zhu, M., Mackintosh, C., Rice, T., Gore, C.,
and Hahn, A., 2007, An Investigation of an Integrated Low-cost
GPS, INS and Magnetometer System for Sport Applications, in
Proceedings of the 20th International Technical Meeting of the
Satellite Division of The Institute of Navigation, Fort Worth,
Texas, USA.
Xiao, B., Zhang, K., Grenfell, R., and Norton, T., 2003,
Handheld GPS: Current Status and Further Developments,
Position Magazine, p. 368-378.
Zhang, K., Grenfell, R., Deakin, R., Li, Y., Zhang, J., Hahn, A.,
Gore, C., and Rice, T., 2003, Towards a Low-Cost, High
Output Rate, Real-Time Rowing Coaching and Training System,
in Proceedings of 16th International Technical Meeting of the
Satellite Division of The Institute of Navigation, Portland,
Oregon, USA, p. 489-498.
Zhang, K., Li, Y., Grenfell, R., Hahn, A., Gore, C., and Rice, T.,
2004, Developing Prototype Multisensor GPS Rower
Monitoring and Coaching Systems for Olympics, in
Proceedings of FIG Working Week 2004, Athens, Greece.
1804