User-friendly Smartphone App for Heart Rate Monitoring in
Sports Endurance Activities
Improving Training Control by Combining New Technologies of
ANT+ and Android
Hans Weghorn
BW Cooperative State University, Kronenstrasse 53A, 70174 Stuttgart, Germany
Keywords: Endurance Training, Heart Rate Monitor, Android, ANTplus.
Abstract: In endurance sports training, heart rate represents a useful proportional measure for the current physical
effort in a workout. Accordingly, during such exercises, it is advantageous to get monitoring information
about the actual heart frequency for having an instrument of controlling the demand level of an activity.
Today, the market offers a broad variety of sports computers, which allow a tracking and display of heart
rate, but their convenience and appropriateness appear rather limited in terms of an efficient use in daily
training. In particular, major restrictions are numerical displays of the in-time measures with tiny letters or
an unfiltered print-out of spurious values, since the recording is not always precise, but sometimes disturbed
from various reasons. Fortunately, upcoming new technologies like programmable smartphone devices and
ANT+ communication standard for sports sensors allow developing new and optimized applications and
systems also for sports purposes. In the work here, a convenient heart rate monitor was developed that aims
at high user-friendliness in combination with elaborated signal-conditioning for preventing any spurious and
misguiding displays. Few simple button presses put the sportsman into the position of performing efficient
activities within the desired endurance training range. As result, a system is described that is feasible for
easy-to-use and efficient sports monitoring, especially during daily workouts.
1 INTRODUCTION
During physical training it is important to know the
actual blood oxygen supply level for categorizing
such activities into the effects of regeneration,
aerobic and anaerobic training. This is required,
because then only dedicate training units will lead in
a controlled and predictable manner to the desired
training goals (Kindermann et al., 1979), for
instance building up speed, over-all endurance or
muscular strength. In practical sports medicine,
blood is extracted from the training sports person
and the analysis of its lactate concentration yields
the state of the oxygen supply, but this method
obviously is too complicate and too expensive for
being continuously used as monitoring instrument in
daily training units.
Fortunately, an alternative measure, which is
applicable at much better convenience, has been
identified since long time: monitoring of the heart
rate (HR) allows a replacement of the direct measure
of the blood oxygen level by a proportional value
(Arts and Kuipers, 1994).
More precisely, as established measure the
percentage value of the actual heart rate in relation
to the maximal possible heart rate (MHR =
maximum heart frequency, also called HR
max
) of the
individual sports person is used. Consequently, this
kind of monitoring approach is commonly applied
for field investigations in medical science (Hoppeler
et al., 1985) and even in sports research (Tabata et
al., 1996). Figure 1 provides a sketch of the main
sports activity levels as these are mapped to different
bands of heart frequency.
For beginners in sports – especially, older ones –
it is established practice that they first build up their
base performance in range RECOM and then after
sufficient advance continue within BE
1
, since
exceeding these levels could even be hazardous for
such untrained people. In general, efficient sports
training requires a certain, well-known mixture of
these activities levels in Figure 1. Else there won’t
be any power advance or even physical deterioration
124
Weghorn H..
User-friendly Smartphone App for Heart Rate Monitoring in Sports Endurance Activities - Improving Training Control by Combining New Technologies of
ANT+ and Android.
DOI: 10.5220/0004613101240133
In Proceedings of the International Congress on Sports Science Research and Technology Support (icSPORTS-2013), pages 124-133
ISBN: 978-989-8565-79-2
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
after a certain point despite continued high training
efforts.
As sample reference, Figure 1 exposes also some
typical ranges for different running activities in the
right column, while the left column in the table
applies to human sports activities in general.
Figure 1: Activity categories as function of HR
max
(the
maximal individual heart frequency).
Electronic consumer market offers today a big
variety of devices for HR monitoring (Figure 2).
This ranges from rather simple and cheap watches to
elaborated sports computers. All of them are tiny,
wearable devices, but their ergonomic use in terms
of user interfacing (UI) appears rather low, when
considering fundamental knowledge and experience
for the research field of human computer interaction
(HCI). Usually the actual heart rate is printed in
letters of limited size and often with limited contrast
on B+W displays. This invokes two fundamental
problems, since this kind of output is
readable at a low speed only
must be categorized by the training persons
themselves (refer to Figure 1)
Certainly, people usually claim that they can easily
remember the desired limits for their training in
terms of lower and upper heart rate. In training
practice, after longer time of demanding activities,
the true experience is quite different: even the
comparison of the simple numbers will get really
slow and causes considerable extra efforts, after, for
instance, attending several hours in a distance run.
Another effect is, that the measuring results from
the heart sensor are sometimes disturbed for
spurious reasons, and displayed totally unfiltered by
the sports watches. In general, systematic errors in
the sensor information have been reported recently
for such sports devices (Weghorn, 2013).
The actual trend of the HR – in particular the
question, whether the heart is getting slowly too high
or too low – is relevant for efficiently controlling an
on-going workout. Despite this interrelation, the HR
trend has to be traced manually by the sports user,
while observing HR number display of the sports
computer and calculating then the slope “by hand”.
Figure 2: Typical sports computers and watches, which
can perform heart rate monitoring as main feature or in
addition to other functionalities.
As positive UI feature, there are also some
modern concepts available; in particular, some
sports computers allow to program the limits for
monitoring the lower and upper HR during the
workouts, and there is even a haptic warning signal
generated, since the device is vibrating, if one of
these limits is exceeded.
In daily use, it often turns out that this feature is
worthless, because the unfiltered HR measures
continuously produce such warning signals.
Furthermore, the warning signal in the sports watch
is often identical for the events of too low and too
high values. This makes it impossible to interpret the
haptic notification correctly without further reading
a message that is displayed in even smaller letters,
since the corresponding printout carries more
information and requires more characters that have
to map on the tiny screen.
From fundamental HCI research it can be
derived, that information perception in computerized
displays can be accelerated by different principles:
User-friendlySmartphoneAppforHeartRateMonitoringinSportsEnduranceActivities-ImprovingTrainingControlby
CombiningNewTechnologiesofANT+andAndroid
125
Using colour as information indicator (Brown,
1999)
Using analogue graphics instead of values that
are in printed in letters
As standard knowledge from the use of technical
systems, it is commonly suggested that an analogue
tachometer is read approx. three times faster than
digitally displayed ones. Although investigations
from the modern application field of automotive do
not proof this number as precise value (Kiefer and
Angell, 1993), they do support the general theory
that analogue displays are perceived at least as fast,
by tendency even faster than digital ones.
As complementing UI aspect, the benefit of
colouring texts is well know and has been
investigated thoroughly during the time, when
coloured computer displays came affordable and
were widely commercialized. Systematic research
has shown that the use of colours increases the
finding speed for a certain passage on a screen
(Carter 1982). The best colouring scheme was found
to use colours that are as distinct as possible for
enabling the fastest perception of different
information contents (Robertson, 1988).
For the user’s convenience, the categorization
and the detection of out-of-band-trends could be
easily performed by the sports computers and data
filtering could prevent the display of erroneous or
misguiding outputs. The calculated results and
trends, and especially the commands, whether the
training should be more challenging or slower for
mapping the desired working range, can be indicated
to the user in a more easily readable manner than
just exposing the plain digits of the current heart
rate. Graphical symbols and/or colouring can be
used for this purpose.
In the following sections, the considerations and
deductions are discussed that led to the design and
construction of an appropriate heart rate monitor that
is intended for sports and health exercises. First
scope was an improved UI that meets standards
according to the established knowledge from the
field of HCI. Second focus was a general
improvement of signal processing, which is
described in the next section, as it depicts the base
for the whole UI concept.
A relevant part of the improvement is also that
the person, who uses this device, is not required to
know all the numbers and relations of sports training
categories, but simple button presses inside the
application allow defining the proper and desired
demand level of the workout. Accordingly, the
system is even more valuable for beginners and in
sports endurance trainings.
2 SIGNAL CONDITIONING OF
HEART RATE DATA
2.1 Low-pass Filtering of Input Stream
For the detection of heart activities, the use of
special chest straps has been established and
commercialized. Such products work really reliable
also for long time of continuous use. Commonly, the
signals are recorded from two skin electrodes across
the lower chest, and are then processed by a tiny
hardware module on the belt that transmits the result
via RF to any data processing sink.
In the experiments here, a comfort chest strap of
the manufacturer Garmin Ltd. was used, its RF
transmission is composed in accordance to the
ANT+ standard (Dynastream, 2011) that is discussed
in the section about technical aspects of this work.
Figure 3: Experiment on the slope of the heart rate. After
slight warming up, three pull-ups were completed while
holding breath during the plotted activity window
(rectangular curve). The heart rate responds by a gradual
increase with delay, while the relaxation phase follows at
an even slower remission speed.
For the fundamental concept of the signal
processing of the raw input data, it is relevant that
the initial sensoring produces four measures per
second, a sampling rate that is typical for such chest
straps. The observation is that sometime these HR
measures appear spuriously distorted, the reasons
have not scientifically been investigated yet, but are
suspected being manifold, e.g. the electrode contacts
are not always accurate due to body movements,
electronic problems, and RF disturbances.
The dedicate reasons may vary, of importance is
mainly, how further processing can overcome such
effects of falsified input data. As signal processing
standard, statistics yields the answer by averaging
consecutive measures as it stands for filtering the
effect of such unavoidable and unpredictable short-
term disturbances.
The properties of such a low-pass filter have to
be aligned to the physiological behaviour of the
icSPORTS2013-InternationalCongressonSportsScienceResearchandTechnologySupport
126
Figure 4: Generation and cumsumption model of the direct and deducted HR measures.
human body in general, since averaging in terms of
any low-pass filtering will suppress the slope of the
output signal and, therefore, may lead to wrong
macroscopic measures, which would automatically
consequence wrong instructions for any on-going
training activities.
Medical research is much older than mobile
computer technology, and hence, base investigations
about the slope of the HR under certain influencing
conditions can be found already from the last
century (Josenhans, 1967). In this extreme, but
indicative experiment, stopping breath followed by
physical effort showed already that the body reaction
in terms of increasing heart beat rate after sudden
load events may be delayed in the order of seconds.
This easily can also be tested with modern
computer technology like in the practical experiment
in Figure 3 (an engineering version of the HR
monitor was used in this test). From both sources, it
can be deducted that averaging HR samples within a
range of few seconds will still yield sufficient results
with adequate slope. Hence, a time slot of 2.5 secs,
which stands for an averaging frame of ten incoming
sensor measures, is applied in the software
construction here. A sliding time-window is used, so
the averigng stage yields also an output rate of 4
values per second, which is synchronized to the
input data stream.
2.2 Feature Extraction
Figure 4 shows the processing chain of the direct
sensoring results as it is used as tier for four output
information streams, which are to be exposed to the
sports user (the corresponding particular UI methods
are discussed in the next section in detail).
In accordance to the different output indicators,
the HR has to be compared to thresholds. For
instance, if a 60-years old person wants to perform
exercises at the effort level of BE
1
, the heart rate
should typically remain within the band from 119 to
136 beats per second. This mapping yields one
important control information for the workout.
Another relevant information is the current trend
of the on-going activity. Is there a trend that the
exercises are running out of the desired band? This
can be detected by differentiating the averaged HR
curve. The derived slope indicates, whether the pulse
is getting too low or too high.
In practice, there will continuously be a slope
that is non-zero, but the relevant matter is, whether
the value is truly running out of band. With human
as part of a control loop, correction indicators have
to be applied carefully, else over-reaction easily may
produce an undesired oscillation in the system. In
general it is difficult to set a precise level of physical
effort, although the body control can improve with
growing experience.
Hence, for out-of-band trends it makes sense to
have two stages of indication: first level is just
exposing the current situation (applied when there is
only slight de-/increase) without any further
instruction. The second is to more clearly indicate
the wrong evelovment and combine it with a
User-friendlySmartphoneAppforHeartRateMonitoringinSportsEnduranceActivities-ImprovingTrainingControlby
CombiningNewTechnologiesofANT+andAndroid
127
Figure 5: Collection of photographs of output screens for different use situations. As seen here, the UI is constructed in a
self-explaining way. For testing of the output colour scheme, the MHR was set to a non-physiological value of 100 bpm for
the right two test images.
particular instruction for the activity like asking to
slow down or speed up.
For the technically and medically interested user
– of course – the unfiltered heart rate may also be of
relevance. In summary, this analysis yields four
output information streams for a display to the user:
the raw data sequence and three filtered and
processed derivatives of it.
3 OPTIMIZATION OF THE USER
INTERFACE
3.1 Visualization of Outputs
The primary goal of the user interface is to provide
an information display that is as clear and as quick
and effortless readable as possible. This display shall
expose the situation and – if required – provide
instructions of how to behave for staying within or –
in case there is a mismatch already – of getting back
to the desired training stress level.
As known from many other applications like, e.g.
car driving and weather stations, an analogue
tachometer is used as main display for the actual
heart rate (Figure 5). The phase vector inside the
tachometer circle is controlled by the averaged HR,
the corresponding indexing hand is drawn
intentionally big for achieving high and quick
readability.
The information, which is to be controlled most
often for an efficient workout, is indicated by colour:
Bright green exposes that the heart rate is within the
desired stress level band. If the HR drops too much,
this colour turns to blue, if it is getting too high, it
turns to red. In this colour indicator, the rule of
readability was infringed intentionally: there are not
only the three, very distinct colours of blue, green
and red in the scheme, but the transition between
these colour states is smoothed.
toolo
w
suffien
t
matchinran
g
e
toohigh
Figure 6: Colour transition scheme for indicating the
match or a gradual mismatch of the heart rate.
Like explained before, the human is a critical
factor in the control loop, because over-reaction
could arise easily by wrong reading or interpretation
of displays: it doesn’t make sense to suddenly turn
the display indicator for the non-mapping heart rate
from bright green to bright red, if the range is just
missed by one single or a few counts, because this
certainly would provoke too strong corrections.
As result, the slope of this indicator is smoothed
by a gradual transition of colours. This scheme was
experimented manually with the display appearance
of the used smartphone, since it turned out that a
straight calculation of ramping up and down of the
RGB base colours looked inconvenient and not
icSPORTS2013-InternationalCongressonSportsScienceResearchandTechnologySupport
128
sufficiently sleek; the best scheme resulted in an
asymmetric number of transition steps (Figure 6).
The raw data value, as it is received unfiltered
from the chest strap sensor, is displayed in big letters
and at high contrast. If the sports user really wants to
know this information from time to time, it is read
easily also from a distance. This appears useful, for
instance, when any type of ergometer or treadmill is
used, and the device has not to be close to the body,
but can be placed in convenient viewing distance
and direction. In such cases it even can be avoided to
turn the head from the normal workout position for
perceiving the information like it is required with
built-in displays of such sports apparatuses, which
are often mounted below chest height.
Just for technical information, a blinking heart
symbol indicates that the RF link to the heart beat
sensor is working correctly (located in Figure 5
below the tachometer centre).
If the system detects that the HR is going to
exceed the defined limits, another UI element is
activated as indicator. A pointer is tilted upwards or
downwards, which stands for the instruction to
speed up or slow down with the on-going effort
respectively (effect is visible in the right two
photographs in Figure 5). Depending on the strength
of the wrong trend, the phasor is tinted more or less
in red tones.
Table 1 summarizes the reading hierarchy, which
stands in accordance to the use frequency of the
information puzzle pieces for dedicatedly controlling
a workout.
Table 1: Use hierarchy of display information.
Indicator
for
UI mode
Frequency
of use
Perception
time
HR inside
range
Colour/analogue ~ 1 / min ≪ 1 sec
HR value Phasor/analogue > 1 / min ~ 1 sec
Instruction Coloured phasor > 1 / min ~ 1 sec
Precise HR Number print rare > 1 sec
3.2 Efficient Input Controls
Totally new in this construction is a convenient
selection of input controls and it is thus the third big
improvement in comparison to commercial heart
rate monitors (Figure 2). As listed in Figure 7, a set
of quick buttons is available on top UI screen,
through which the user can set the monitoring level
according to Figure 1. For this, it is required that the
MHR is configured once in the application
(accessible through standard settings menu like is
implemented in all smartphone apps).
This MHR parameter is critical, best is, if it is
determined under supervision of an experienced
doctor. There are many, different formulas in
various sports discussions panels found on the
Internet as also in various research papers. The two
described main methods, which yield slightly
different results, are based on the age of the person.
No valid scientific foundation for these calculation
rules was found in literature, and hence the sports
user is not offered in the heart monitor to enter the
age, since the calculation of a wrong MHR could be
risky or at least lead to inefficient training ranges.
The quick buttons for the effort level simplify the
control considerably. According to the invoked
function behind, it is not required, that a sports user
does know all the relations between training goals
and heart rates, and the system also releases the
sports user from continuously validating in his head,
whether the HR maps like desired. Besides the
standard effort levels, there is also a button for
activating a more dedicate trainings plan. This
initiates a well defined sequence of varying effort
levels; for instance, it may be programmed that the
sports user first works ten minutes in RECOM, then
10 minutes in BE1, then 10 minutes in BE2 and
before terminating the workout, again 5 minutes in
BE1. Of course, such a plan has to be entered with
all its stages into the system, but this method is
required for advanced sports training anyway.
Figure 7: Quick buttons design for setting the desired
training effort level – collection of active and in-active
screen buttons (refer to use cases in Figure 5).
There is one button for activating the RF link to
the heart belt, which is of technical function. If the
air link is active, a beating heart symbol inside the
HR tachometer indicates the working connection
(Figure 5).
There are further buttons, which can be relevant
for the trainings: one for starting, and a second for
pausing and stopping a workout (low part of the UI
screens in Figure 5). The starting button stands for a
recording of the heart rate samples. It has got a
double function as it also starts a programmed
training sequence as described before, if this mode is
activated.
At the moment, the recording is stored to a
XML-coded file on the multi-media card of the
phone. From there, it can be directly downloaded
User-friendlySmartphoneAppforHeartRateMonitoringinSportsEnduranceActivities-ImprovingTrainingControlby
CombiningNewTechnologiesofANT+andAndroid
129
Figure 8: The XML recording of the HR monitor can be imported through a USB cable from the smartphone by a few
mouse clicks directly into any sports analysis software on a personal computer (the “Garmin Ltd. Training Center” was
used here). In this experiment, a 3 kms inline skating lap was completed, main output is the analysis oh HR plots (red curve
in the lowest sub-window). As feature that complements the HR tracking in outdoor sports activities, GPS positions were
recorded during the test also: this was working with sufficient quality despite the fact, thar the HR device was carried inside
a rucksack on the back of the sports person.
into any compatible sports evaluation software,
which is running on a personal computer (e. g. the
Garmin training center, or similar others). On the
personal computer, the records can then be analyzed
and archived for a long-term planning and trace of
workout activities. If the GPS signal is available, the
heart rate records are also tagged with GPS position
information (result of a skating test lap visible in
Figure 8). Since the XML format is not space
saving, the GPS tagging is performed at a low rate (4
samples per minute) to keep the total files size
within reasonable limits.
All control buttons are designed as “self-
expressive”, which is achieved by using standard
nomenclature and commonly used standard symbols.
4 TECHNICAL ASPECTS AND
REALIZATION
Similar to the evolvement of mobile phones in their
first generations, sports watches and computers are
closed technical systems, regardless whether they
are simple or elaborated. Neither their functionality,
nor their software can be modified. For smartphones,
it is meanwhile well established that other people
than just the constructors of these devices can bring
own software applications on these units or can even
extend hardware through standard interfaces. For the
experiments here a smartphone with the “open”
operating system Android was chosen (Collins et al.,
2011), since there is lots of support and introductory
material available in terms of books and on the
Internet in documentation and developers panels.
Industry for sports computers has agreed several
years ago on a new RF communication standard,
which enables the interoperability of devices from
different vendors. In the past, sports device
constructors used their proprietary air link solutions,
nowadays the so-called ANT+ standard
(Dynastream, 2011) stands for an efficient data
transfer from typical sensors like heart belts and
tread or speed sensors for running and cycling. Like
Bluetooth and WLAN, ANT+ uses the ISM
frequency band, but at much lower energy
consumption, which makes it possible that a heart
frequency belt can run several days from the energy
of one single lithium battery cell.
Since the corresponding working consortium for
ANT+ is interested in spreading the concept, it
icSPORTS2013-InternationalCongressonSportsScienceResearchandTechnologySupport
130
claims that there is a long list of smartphones, which
do supply this communication link as well. In the
end, the number of phones with built-in ANT+
interface certainly is growing, but nevertheless, it is
not really broadly available, if all the available
smartphones products are considered.
Figure 9: Upper arm pocket bag as it is part in the delivery
package of the outdoor smartphone. This wearing solution
does appear neither comfortable nor useful for any kind of
sports activities.
At the time, when the investigations in this
research were started, there were few and expensive
ANT+ phones only. In this project, a Sony Xperia
Active was selected as target device, it has got an
older version of the Android operating system, but it
was constructed especially for outdoor and sports
use and it comes with a special pocket for wearing it
at the forearm or upper arm during such activities
(Fig. 8). There is good software support for
developers from the phone manufacturer, but the
ANT+ application software technology is complicate
anyway, and it is proprietary.
With newer versions of smartphone operating
systems, there may be a built-in standard software
interfacing to ANT+. This concept theoretically will
allow transferring applications, like the one in this
project, more easily to devices of other brands. A
possible greater target market for the heart rate
monitor is moreover supported, because meanwhile
there are heart frequency belts commercially
available, which interface through Bluetooth instead
of ANT+. Bluetooth certainly maps almost all new
smartphones, but some doubts arise, whether these
belts can be as power-saving and will have by that
appropriate operational times.
Since the research question in this work is not of
commercial interest, the development will continue
also together with other related investigations on
sports sensors on this given hardware platform,
despite the fact that the hardware/software approach
can be considered already as overtaken by new
technology generations on the market. On the other
hand, the used phone still is sold on the market and
advertised by its vendor as modern outdoor device.
5 APPLICATION SCENARIOS
Since the heart rate monitor is a versatile tool and
not specific to any particular kind of sports, it can be
used broadly. Especially, endurance exercises and
training – and by that building up general physical
fitness in sports or health – is the best field for its
application. The smartphone can be mounted on a
bicycle or on any type of sports machine in a way
that it is easily visible by the user without unnatural
movements of the head. Although training utilities
like, e. g., treadmills and cycloergometers usually
have built-in heart rate monitoring as well, these
devices do often not easily or even do not at all work
with custom heart frequency belts. At least, there is a
procedure for registering an individual one and –
furthermore – there may remain hygienic concerns
when sharing such a skin touching belt with other
people. After the workout it is often not possible to
get access to the records of the training inside the
computer of the sports apparatus.
Hence, using an own heart rate monitor also for
such gymnastic machines provides in total several
advantages, namely the opportunity for a natural and
comfortable use during the workout, hygienic
advantages and the possibility for easily preserving
the records of the workout for later analysis and
planning.
6 DISCUSSION OF CONCEPTS
AND THEIR CONSTRAINTS
From well-known fundamentals of HCI research, the
described approach of constructing a user-friendly
heart rate monitor arises partially as straight-forward
design. Of course, certain detail questions (e. g.
transition scheme for colour indicators, averaging
duration of raw data input) had to be researched in
literature and complemented by own practical
experiments.
Although the system appears aaplicable for its
intended purpose, the usability has to be proven yet
by scientific methods. This is planned during the
upcoming summer season mainly in outdoor
User-friendlySmartphoneAppforHeartRateMonitoringinSportsEnduranceActivities-ImprovingTrainingControlby
CombiningNewTechnologiesofANT+andAndroid
131
activities. The system has got also some limitations,
which arise from its technical construction. For
instance – due to the RF communication inside the
ISM frequency band – it is unusable for swimming
or under/in-water activities. Since operating on a
smartphone, which in the end has got some physical
dimensions and weight, the system will also not be
helpful for sports competitions. For this, much
smaller and less cramping devices similar to small
watches are required.
During running, cross-country skiing and all
other comparable activities an extra carrier pouch
has to be used, which is also part of the delivery
package of the used phone (Figure 9). Considering
the size and weight of this solution, this obviously
cannot be accounted as appropriate solution.
The preferable way for such kind of sports
activities would be to run the smartphone as
intelligent relay, which is carried, e. g., in a belt or a
bag close to the body, and to use in addition a
watch-like system worn at the wrist as remote
display in terms of software layers of the HR
monitor.
Figure 10: Remote wrist display for Android smartphone.
This device could perfectly complement the heart rate
monitoring UI for endurance sports like distance running.
Such hardware indeed is available for the phone
brand, on which the system was implemented for
this work (Figure 10). Using this remote display
extension would be even more comfortable than
sports watches, because it is lighter and has got a
colour display and a comparable operational time.
At current project state, this concept extension was
just started, additional fundamental work will be
required especially on the UI aspects, since the
display size of the wrist device is much smaller and
possible input actions are simplified in comparison
to a full smartphone screen. Furthermore, the
software needs to be specially designed for low-
power consumption; optimization for this aspect was
not regarded in the prototypes that have been
developed so far.
The feasibility of this relaying smartphone
system was also tested already in this research, but
so far only with a few singular experiments. In these
tests, the smartphone with the activated HR monitor
was carried inside a rucksack on the back, while
performing running and inline skating laps (skating
test sample in Figure 8). In parallel, the lap routing
was recorded with other commercial tracking
devices; a comparison of results showed that the
difference between the HR distance measure and the
other systems was not more than 1%, which appears
well acceptable, since the precise Geo tracking is
none of the primary scopes of the HR monitor.
Considering sports physiology, there arises a
general question with the concept of linking
workouts to the MHR. Especially in endurance
training, different energy reservoirs in the human
body are used in a sequence. During the first
minutes, phosphate storage is used, which is located
inside the muscles, and low oxygen is required for
burning this. Consequently, the heart rate is lowered
in the beginning of sports activities. After this – for a
phase of approx. 1.5 hours – the body uses
carbohydrate burning, which maps in general well to
the HR levels in Figure 1. Afterwards, the body tries
to supply itself by “fat” burning (i. e. a conversion of
fat reservoirs into carbohydrate, which is then used
as energy supply for the muscles), if the person is
used to it. In this phase, the heart rate starts
increasing considerably.
But even in the phase of carbohydrate burning at
steady aerobic level, there exists the so-called effect
of “cardiodrift” (Dawson et al., 2005). It stands for a
continuous, slight increase of HR, which occurs
despite a perfectly balanced demand level already
for workouts less than an hour. Cardiodrift is not
fully understood yet, there even exists some
controversial discussion since longer time about its
origin, but the effect itself is not under question and
hence, it would to be considered in an accurate
training regulation as well. The control loops in the
developed HR monitor are not prepared yet for
compensating this effect, at the current state of
research reports it wouldn’t anyway be possible to
identify a general mathematical rule for
continuously adopting the HR band limits during an
on-going workout.
This all implies that HR monitoring doesn’t work
well as effort indicator for short workouts, and
appears only reasonable with a programmed training
plan for longer workouts. Without detail knowledge
about these relations or planning support and
icSPORTS2013-InternationalCongressonSportsScienceResearchandTechnologySupport
132
guidance from appropriate experts, HR-controlled
training units may have only limited effects, when
the workouts are too short or too long.
7 CONCLUSIONS AND
OUTLOOK
The concept of a user-friendly heart rate monitor has
been researched scientifically, and it has been
realized on base of the corresponding findings
technically. In this system, coloured and analogue
display indicators allow perceiving the most relevant
heart monitoring information with quick and
effortless glances in less than a second. The training
range can be conveniently set by one single button
touch on top-level screen.
This work represents a knowledge fusion from
the areas of human-computer interaction (HCI),
mobile computing, human physiology and sports. As
result the system is useful for controlled sports
training as also for health exercises in a broad
application range, where heart rate stands for the
effort level. It can be used indoor and outdoor for
general gym exercises, strength training, walking,
running, cycling and skiing.
Best comfort certainly is reached, when the heart
monitor is not carried close to the body, but mounted
in best viewing distance and direction. The latter
would be achieved easily by commonly available
smartphone holders for bicycles and automotive.
For outdoor sports without machines, certainly
the current system is sensed as not being optimal due
to its size and weight; some additional development
would be required to improve it. Although there was
positive tendency in the feedback of first friendly
user tests, there is no scientifically-based proof
behind this assessment about the usability yet. This
still has to be validated and the corresponding
findings certainly will lead to additional detail
improvements.
Already now – in comparison to consumer
market devices – the heart rate monitor, which was
developed here, is funded on appropriate and
modern HCI concepts and it will by that ease the
control of sports activities for achieving in the end
the desired training results more efficiently.
REFERENCES
Arts, F. J., Kuipers H., 1994. The relation between power
output, oxygen uptake and heart rate in male athletes.
Int. J. of Sports Medicine, Vol. 15, No. 5, pp228-231.
Brown, M. B., 1999. Human-Computer Interface Design
Guidelines. Intellect Books Ltd., Exeter, UK, Chapter
Four “Color”, pp 66-79.
Carter, R. C., 1982. Visual search with color. J. of
Experimental Psychology. Human Perception and
Performance, Vol. 8(1), pp 127-136.
Collins, C., Galpin M., Kaeppler M., 2011. Android in
Practice. Manning Publications, Westampton, USA.
Dawson, E. A., Shave, R., George, K., Whyte, G., Ball,
D., Gaze, D., Collinson, P., 2005. Cardiac drift during
prolonged exercise with echocardiographic evidence
of reducted diastolic function of the heart. European J.
of Applied Psychology, Vol. 94(3), pp 305-309.
Dynastream Innovations Inc., 2011. ANT message
protocol and usage. Sourced from http://thisisant.com,
Rev. 4.5.
Josenhans, W. T., 1967. Breath holding effects on ULF
displacement ballistocardiography. Bibl Cardiol., Vol.
19, pp 49-62.
Hoppeler, H., Howald, H., Conley, K., Lindstedt, S. L.,
Claassen, H., Vock, P., Weibel, E. R., 1985.
Endurance training in humans: aerobic capacity and
structure of skeletal muscle. J. of Applied Physiology,
Vol. 59, No. 2, pp 320-327.
Kiefer, R. J., Angell, L. S., 1993. A comparison of the
effects of an analog versus digital speedometer on
driver performance in a task environment similar to
driving. J. of Vis. Veh., Vol. 4, pp 283-290.
Kindermann, W., Simon, G., Keul, J., 1979. The
Significance of the Aerobic-anaerobic Transition for
the Determination of Work Load Intensities During
Endurance Training. Eur. J. Appl. Physiol., Vol. 42,
pp 25-34.
Robertson, P. K., 1988. Visualizing color gamuts: a user
interface for the effective use of perceptual color
spaces in data displays. IEEE Computer Graphics and
Applications, Vol. 8(5), pp 50-64.
Tabata, I., Nishimura, K., Kouzaki, M., Hirai, Y., Ogita,
F., Miyachi, M., Yamamoto, K., 1996. Effects of
moderate-intensity endurance and high-intensity
intermittent training on anaerobic capacity and
VO2max. Medicine & Science in Sports & Exercise,
Vol. 28(10), pp 1327-1330.
Weghorn, H., 2013. Applying mobile phone technology
for making health and rehabilitation monitoring more
affordable. Biosignals and Biorobotics Conference
(BRC), 2013 ISSNIP, Rio de Janeiro, Brasil, pp 1-5.
User-friendlySmartphoneAppforHeartRateMonitoringinSportsEnduranceActivities-ImprovingTrainingControlby
CombiningNewTechnologiesofANT+andAndroid
133
