Note: Descriptions are shown in the official language in which they were submitted.
CA 02860129 2014-08-20
1
SYSTEM, DEVICE AND METHOD FOR QUANTIFYING MOTION
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The
present application claims priority on US 61/867,703 filed on
August 20, 2013, that is hereby incorporated by reference in its entirety.
FIELD OF THE DISCLOSURE
[0002] The
present disclosure relates generally to systems, devices and
methods for motion capture, motion tracking, motion detection and motion
quantification, especially for sport applications. More specifically, the
present
disclosure relates to the automatic identification, measurement and
quantification
of a player's individual performance, such as a hockey player.
BACKGROUND OF THE DISCLOSURE
[0003] Sport
practitioners are eager to acquire motion metrics indicative of
their training or gaming performance, for primarily performance improvement
purpose, and secondarily for comparing their performance to those of their
peers
or of top professional athletes.
[0004] Sensor
based motion tracking systems are limited in the complexity
of gestures or movements that are able to quantify. Inertial sensor products,
also
known as IMU (inertial measurement units), are confined to motion tracking in
sports where the predictability of actions is easier. Sport applications using
IMUs
remain captive of repetitive movements and simple gestures such as stride,
walking or running. A stride, using a single motion sensor positioned at the
foot,
is probably the simplest movement to quantify.
[0005] Current
sensor based motion tracking systems suffer from several
limitations. No such motion tracking system today is able to detect,
discriminate
and quantify complex gestures in a non-choreographed motion sequence. As an
example, IMUs wearable on a golf glove which can detect the acceleration,
velocity, tempo, position and posture of the sensor and estimate the position
and
movement of the golf club, are among the most advanced movement
quantification systems on the market today. To use such a golf swing
quantifier,
CA 02860129 2014-08-20
2
the player must typically manually activate and/or configure a sensor prior to
a
swing, and again at the end of the swing movement, in order to specify the
movement sequence to be quantified. Manual activation of the sensor prior to
each movement is impractical in most sports where movements are not
choreographed, such as racket sports and team sports.
[0006] Moreover, currently available IMU based motion tracking systems
are limited by at least one of the two following elements: (a) they lack the
capability of blind identification of motion patterns and/or (b) they lack the
capability of blindly identify complex motions. In (a), by "blind" we refer to
the
non-choreographed nature of motion executed in most sports where the player
does not "inform" the sensor of its intention to make a specific move. As an
illustration, IMU-based, golf movement quantification products cannot
discriminate a "golf drive" motion from a "golf putting" motion. The user must
instruct the sensor of its intentions prior to each specific move. In (b), by
"complex" we refer to motion that requires more than acceleration measurements
to be detected and identified. In other commercial products, run movement
quantifiers use either simple accelerometers or pressure sensors to detect
steps,
and then compute the number of steps per second to evaluate the running
speed. Jump movement quantifiers for basketball use shoe-based pressure
sensors to estimate jump height. Both products focus on simple motion, the
"steps", that are easy to detect and quantify by identifying the zero-velocity
states, with accelerometers, or the zero pressure states with the pressure
sensors. All currently available IMU based products cannot identify more
complex motion hidden in motion noise. Referring to the examples mentioned
above, a run movement quantifier could not detect a ball kick in a soccer
game.
A jump quantifier for basketball could not discriminate a three point shot
from a
dunk shot maneuver.
[0007] Consequently, no system of the prior art provides motion tracking
system based on inertial sensors optimized for complex gestures. No system of
the prior art provides motion tracking system based on inertial sensors which
detect, discriminate and quantify complex gestures in a non-choreographed
CA 02860129 2014-08-20
=
3
motion sequence. No system of the prior art provides motion tracking system
based on inertial sensors which can automatically operate without being
activated. No prior art system provides a motion tracking system based on
inertial sensors that can be operated by a player for long period of time
without
the need for external systems.
[0008] Integration of motion sensors on sport equipment
requires an
important miniaturization effort to minimize sensors' weight and avoid
obstruction
so that ideally the sensors become unnoticed and do not affect movement in any
way. Manufacturing cost considerations also calls for a diminution in the
number
of electronic components involved in the movement sensors.
SUMMARY OF THE DISCLOSURE
[0009] It would thus be highly desirable to be provided with an
apparatus
or a method that would at least partially solve one of the problems previously
mentioned or that would be an alternative to the existing technologies.
[0010] According to one aspect, there is provided device for
motion
identification, the device comprising: an enclosure; and a plurality of
sensors
being provided with the enclosure and configured to measure acceleration in
three axes and angular motion in three axes.
[0011] According to another aspect, there is provided a device
for motion
identification and at least one external sensor being external to the device,
and
being configured to measure at least one of vibration, acceleration, rotation,
magnetic field, temperature, humidity, flexion, bend, orientation, distance to
an
object, sound, image, heart-beat, blood, wind pressure at the external sensor,
wherein the controller is further configured to receive at least one
measurement
from the at least one external sensor, and wherein characterizing the user-
executed movement is further based on the at least one measurement received
from the at least one external sensor.
[0012] According to yet another aspect, therein provided a
method for
motion identification, the method comprising: receiving one or more
CA 02860129 2014-08-20
4
measurements taken by a plurality of sensors during an user-executed
movement, the measurements being representative of the user-executed
movement and comprising acceleration measurements in three axes in a first
acceleration range, acceleration measurements in three axis in a second
acceleration range, and angular motion measurements in three axes, and
characterizing the user-executed movement based on the received one or more
measurements.
BRIEF DESCRIPTION OF THE DRAWINGS
[0013] The following drawings represent examples that are presented in a
non-limitative manner.
[0014] FIG. 1 is a schematic projected view showing the components of
the inertial sensor unit.
[0015] FIG. 2 is a schematic projected view of the enclosure part of the
inertial sensor unit with stopper parts.
[0016] FIG 3A-B-C are graphs depicting the motion patterns associated
with a hockey slap shot motion.
[0017] FIG. 4A-B are flowcharts exemplifying the functioning of the motion
quantifying algorithm.
[0018] FIG. 5 is an organisation chart showing the network components
used by the motion quantification system and method when transmitting, using
and managing performance data.
[0019] FIG 6A-D are representations of interfaces using a hand held
device, for the presentation of user profiles, performance statistics,
communication of metrics and comparison of quantified motion data for a single
user and several users.
DETAILED DESCRIPTION OF THE DISCLOSURE
[0020] The following examples are presented in a non-limitative manner.
CA 02860129 2014-08-20
[0021] "stick" as used herein refers to any sport instrument such as a
stick
or bat used to contact an object such as a ball, including without limitation
a
hockey stick, a baseball bat, a cricket bat, a golf club, a tennis racket, or
a
badminton racket.
[0022] "puck" or "ball" as used herein refers indistinctively to any
object
that is intended to interact with a stick during sport, such as without
limitation, a
hockey puck, a golf ball, a tennis ball, another hockey stick.
[0023] According to various exemplary embodiments, there is provided a
device for motion identification and quantification, comprising: An inertial
sensor
to provide motion data, comprising a plurality of inertial sensors including
at least
one but typically three accelerometers for measuring acceleration in three
axis, at
least one but typically three gyroscopes for measuring angular motion in three
axis, a microcontroller executing algorithms to detect, interpret and quantify
motion data, a memory means to store motion data and quantified performance
data, a telecommunication means to connect to a remote computer, wherein the
inertial sensor measures high acceleration (ex: greater than or equal to about
10G, greater than or equal to about 15G, between about 10G to about 20G, or
tens of G for measuring high impacts) and small acceleration (ex: less than
about
3G, less than about 10G, or less than about 15G), at different frame rates
including a frame rate higher than 1000 measurements per second and a frame
rate of less than 1000 measurements per second, wherein the device for motion
identification and quantification is located on a single measurement point
located
on a stick or on a user.
[0024] For example, the inertial sensor comprises a magnetometer for
measuring the strengths and direction of magnetic fields.
[0025] For example, the device is positioned inside the hollow section of
a
stick handle.
[0026] For example, the device is positioned near the extremity of a stick
handle.
CA 02860129 2014-08-20
6
[0027] The device comprises an enclosure adapted to fit inside the hollow
section of a stick, using stopper parts located at the bottom and top sections
of
said enclosure, each stopper part consisting of several tongues located on at
least two sides of the stopper part.
[0028] For example, the device is by default in continuous operation and
activates itself automatically by motion, wherein one sensor of either said
accelerometer group or gyroscope group is kept powered in wake mode at all
time while the inertial sensor is otherwise turned off, and wherein said one
sensor can detect a vertical, lateral or rotational motion, which triggers
powering
on of the device.
[0029] For example, the device recognises a stick among a plurality of
sticks identified with barcodes or RFID tags, wherein the device is provided
with
a barcode or RFID tag reading means, and wherein the sensor unit is affixed on
the exterior surface of the stick, or wear by user on a glove or on a wrist
[0030] For example, the device automatically recognises when user
executes a motion.
[0031] For example, the device automatically recognises when user hits
an object with the stick, by detecting the vibrations and impact associated
with
said hit.
[0032] According to various exemplary embodiments, there is provided a
system for motion identification and quantification, comprising: an inertial
sensor
to provide motion data, comprising a plurality of inertial sensors including
at least
one but typically three accelerometers for measuring acceleration in three
axis, at
least one but typically three gyroscopes for measuring angular motion in three
axis, a microcontroller executing algorithms to detect, interpret and quantify
motion data, a memory means to store motion data and quantified performance
data,a telecommunication means to connect to a remote computer, a remote
computer for presenting and communicating motion data and quantified
performance data,at least one external sensor connected or communicating with
said inertial sensor, from a group of sensors comprising sensors for sensing
CA 02860129 2014-08-20
7
vibration, acceleration, rotation, magnetic field, temperature, humidity,
flexion,
bend, orientation (such as magnetometer), distance to an object, sound, image,
heart-beat, blood, wind pressure, wherein raw sensing data from said external
sensor is processed by said microcontroller.
[0033] For example, external sensor is positioned on the stick or on a
user.
[0034] For example, the external sensor is a heterogeneous or
homogeneous combination of several such external sensors.
[0035] For example, the said external sensor is a bending sensor to
recognise mechanical bend of the appropriate part, affixed on or moulded in
the
blade part or the shaft part of a stick.
[0036] According to various exemplary embodiments, there is provided a
method for installing and using a sensor device in the hollow section of a
stick
handle such as a hockey stick, the sensor device having an enclosure adapted
to fit inside the hollow section of a stick, using stopper parts located at
the bottom
and top sections of said enclosure, each stopper part consisting of several
tongues located on at least two sides of the stopper part, a comprising the
steps
of removing the cap that closes the hollow section at the end of the stick,
placing
the sensor unit in the hollow section of the stick, providing a progressive
pressure
downward on the sensor unit, until only the top part of enclosure 90 covers
the
top part of the stick handle, powering on the sensor device by a motion of the
stick.
[0037] According to various exemplary embodiments, there is provided a
method for identifying and quantification motion, using a sensor unit
comprising
an inertial sensor for providing acceleration and rotation data in three axis,
a
microprocessor executing algorithms for identifying and quantifying motion and
memory means, located on a user or on a stick, comprising the steps of:
providing raw motion data using the inertial sensor mounted on a stick or on a
user, said raw motion data comprising acceleration in three axis and gravity
data
including angular speed variation, roll, pitch and yaw, for a given duration,
CA 02860129 2014-08-20
8
detecting by correlation, from said raw motion data, the presence of motion
patterns similar to motion patterns stored on a memory located on said
inertial
sensor, which are representative of the motion patterns associated with
movements typically performed during practice of a given sport, using
algorithms
executed on a microcontroller on said inertial sensor, identifying by
correlation,
from the detected motion patterns, the type of movements performed by user,
among a library of typical movements associated with a given sport stored on
said memory, using algorithms executed on said microcontroller, confirming if
a
motion pattern qualifies or not as a typical movement, using algorithms
executed
on said microcontroller and, if a motion pattern qualifies, quantifying said
motion,
and providing quantified motion data, using raw sensing data on said memory
associated with said motion processed by a second algorithms executed on said
microcontroller, storing the quantified motion data and its associated raw in
said
memory, wherein quantified motion data provides information on a motion
including its duration, its speed and angle amplitude.
[0038] For example, regarding the motion of a stick, the second algorithm
computes an estimation of the maximum speed of an object hit by said stick
that
is representative of the real speed of said object immediately after impact,
using
raw data provided by said inertial sensor and quantified motion data.
[0039] For example, said quantified motion data is correlated by a second
algorithm executed by microcontroller using at least one value from a group of
motion metrics including "swing amplitude", "swing duration", "stick speed",
"stick
acceleration", "wrist effect", "detailed flex analysis", "motion amplitude"
and
"motion behaviour".
[0040] For example, said quantified motion data is correlated by second
algorithm executed by microcontroller using at least one user centric data
from a
group of user data including "user age", "user gender", "user location", "user
height", "user weight", "stick length", "stick model" and "stick flex".
[0041] For example, the algorithm, while identifying a motion pattern,
discriminates and quantifies a motion pattern among motions that share same or
CA 02860129 2014-08-20
9
equivalent motion patterns stored on said memory, using a larger set of motion
metrics that are pre-established for each motion pattern.
[0042] For example, the sensor unit comprise a communication means
used to transmit quantified motion data to the remote computer.
[0043] For example, the sensor unit is used in recording mode while
specific motions are executed by user.
[0044] According to various exemplary embodiments, there is provided a
method for identifying and quantification motion, using a sensing unit
comprising
an inertial sensor for providing acceleration and rotation data in three axis,
a
microprocessor executing algorithms for identifying and quantifying motion and
memory means, located on a user or on a stick, comprising the steps of:
automatic activation by a motion of the sensing unit, starting motion
detection,
wherein said sensor unit is initialized and starts real-time monitoring and
analysis
of sensing data from the inertial sensor, starting a potential motion event,
when
the algorithm detects and identifies a motion pattern associated to a
backswing
motion, starting a potential motion acceleration event, when the algorithm
detects
and identifies a motion pattern that is greater in value than a predetermined
lowest threshold and lower in value than a higher threshold, starting a
potential
motion continuation event, when the algorithm detects and identifies a motion
pattern associated to a potential downswing motion, starting potential impact
detection metric, by which the algorithm monitors the variation of the sensing
data over an impact detection period, and log the event as a potential impact
when the variation reaches a predetermined level, starting impact detection
metric, by which the algorithm detects and identifies an impact event
following a
downswing motion within a predetermined time period, starting shot detection,
by
which the algorithm confirms an impact event, and quantifies the motion
metrics
associated with the event.
[0045] For example, sensing data from the inertial sensor provides metric
value expressed in m*rad/s3 as outputs.
CA 02860129 2014-08-20
[0046] According to various exemplary embodiments, there is provided a
system for motion identification and quantification, comprising: a sensing
unit
comprising an inertial sensor for providing acceleration and rotation data in
three
axis, a microprocessor executing algorithms for identifying and quantifying
motion, memory means hosting a database, and communication means, located
on a user or on a stick, a remote computer synchronized with the sensing unit
through the communication means, for updating data and usage statistics, an
external computer with a database, connecting to the remote computer using a
wired or wireless connection, an online platform accessible form the remote
computer and the external computer, wherein sensing data from the inertial
sensor and quantified motion data provided by the algorithm are transmitted
from
the sensing unit to the remote computer using the communication means.
[0047] For example, the remote computer connected to the online platform
enables user of the remote computer to present and communicate its sport
performance to other users and to compare its sport performance to those of
other users.
[0048] For example, wherein the system comprises hand held devices
connected to the external computer or to the remote computer that allows third
parties to visualize, present, share and compare quantified motion information
and user performance data.
[0049] For example, according to devices of the present disclosure, the
plurality
of sensors comprises three accelerometers for measuring acceleration in three
axes in the first acceleration range and in the second acceleration range.
[0050] For example, according to devices of the present disclosure, the
plurality
of sensors comprises at least three gyroscopes for measuring the angular
motion
in three axes.
[0051] For example, according to devices of the present disclosure, the
plurality
of sensor comprises a first accelerometer group of at least three
accelerometers
for measuring acceleration in three axis in the first acceleration range and a
CA 02860129 2014-08-20
11
second accelerometer group of at least three accelerometers for measuring
acceleration in three axis in the second acceleration range.
[0052] For example, according to devices of the present disclosure, the
plurality
of sensors comprises at least three gyroscopes for measuring the angular
motion
in three axes.
[0053] For example, according to devices of the present disclosure, the first
acceleration range is between about OG to about 10G and wherein the second
acceleration range is from at least 10G and above.
[0054] For example, according to devices of the present disclosure, the second
acceleration range is from about 10G to about 100G.
[0055] For example, according to devices of the present disclosure, the
plurality
of sensors comprises a plurality of accelerometers, a first subset of the
accelerometers measuring at a high measurement rate of more than 1000
measurements per second and a second subset of the accelerometers
measuring at a low measurement rate of less than 1000 measurements per
second.
[0056] For example, according to devices of the present disclosure, the
plurality
of sensors comprises a plurality of gyroscopes, a first subset of the
gyroscopes
measuring at a high measurement rate of more than 1000 measurements per
second and a second subset of the gyroscopes measuring at a low measurement
rate of less than 1000 measurements per second.
[0057] For example, according to devices of the present disclosure, the
plurality
of sensor comprises a magnetometer for measuring the strength and direction of
magnetic fields.
[0058] For example, according to devices of the present disclosure, the
plurality
of sensors comprises a plurality of magnetometers, a first subset of the
magnetometers measuring at a high measurement rate of more than 1000
measurements per second and a second subset of the magnetometers
CA 02860129 2014-08-20
12
measuring at a low measurement rate of less than 1000 measurements per
second.
[0059] For example, according to devices of the present disclosure, the
plurality
of sensors are located within the enclosure in close proximity of one another.
[0060] For example, according to devices of the present disclosure, the
enclosure
is adapted to be positioned in the hollow section of a stick handle.
[0061] For example, according to devices of the present disclosure, the
enclosure
is adapted to be positioned proximate the extremity of a stick handle.
[0062] For example, according to devices of the present disclosure, the
enclosure
comprises one or more stopper parts each having one or more tongues for
gripping the interior surface of the stick handle.
[0063] For example, according to devices of the present disclosure, the
plurality
of sensors are operable between a sleep state and a wake state, wherein at
least
one of the sensors remains in a wake state while the remainder of the sensors
are operating in a sleep state, and wherein detection of a motion by the at
least
one sensor in the wake state triggers activating at least one of the remainder
of
the sensors to the wake state.
[0064] For example, devices of the present disclosure further comprise a
memory
to store measurements taken by the plurality of sensors; and a communication
module for transmitting the stored measurements to an external device.
[0065] For example, devices of the present disclosure further comprise a
controller configured for receiving one or more measurements taken by the
plurality of sensors during an user-executed movement; and characterizing the
user-executed movement based on the received one or more measurements.
[0066] For example, devices of the present disclosure further comprise a data
storage device having stored thereon a plurality of sets of predetermined
metrics;
and wherein characterizing the received measurements comprises correlating
the received measurements with the sets of predetermined metrics; and
CA 02860129 2014-08-20
13
detecting whether the received measurements substantially matches one of the
sets of predetermined metrics.
[0067] For example, according to devices of the present disclosure, the
controller
is further configured for when the received measurements substantially matches
one of the sets of predetermined metrics, identifying the type of the user-
executed movement.
[0068] For example, according to devices of the present disclosure, the
controller
is further configured for quantifying the user-executed movement based on the
received measurements.
[0069] For example, according to devices of the present disclosure,
quantifying
the user-executed movement comprises quantifying a duration, speed and angle
of the user-executed movement.
[0070] For example, according to devices of the present disclosure, the angle
of
the user-executed movement comprises an angle in a first plane and an angle in
a second plane.
[0071] For example, devices of the present disclosure further comprise an
identification tag reader.
[0072] For example, a system for motion identification further comprises a
devices as disclosed herein, at least one external sensor being external to
the
device, and being configured to measure at least one of vibration,
acceleration,
rotation, magnetic field, temperature, humidity, flexion, bend, orientation,
distance to an object, sound, image, heart-beat, blood, wind pressure at the
external sensor; wherein the controller is further configured to receive at
least
one measurement from the at least one external sensor; and wherein
characterizing the user-executed movement is further based on the at least one
measurement received from the at least one external sensor.
[0073] For example, according to systems of the present disclosure, the
external
sensor is positioned on one of a user and a stick used during the user-
executed
movement.
CA 02860129 2014-08-20
14
[0074] For example, according to systems of the present disclosure, the
external
sensor is a bending sensor for measuring a mechanical bend of one of a blade
part of a stick and a shaft part of a stick.
[0075] For example, according to systems of the present disclosure, the
external
sensor takes measurements at a high measurement rate or more than 1000
measurements per second and at a low measurement rate of less than 1000
measurements per second.
[0076] For example, according to methods of the present disclosure,
characterizing the received measurements comprises correlating the received
measurements with a plurality of sets of predetermined metrics; and detecting
whether the received measurements substantially matches one of the sets of
predetermined metrics.
[0077] For example, methods of the present disclosure further comprise when
the
received measurements substantially matches one of the sets of predetermined
metrics, identifying a type of the user-executed movement.
[0078] For example, methods of the present disclosure further comprise
quantifying the user-executed movement based on the received measurements.
[0079] For example, according to methods of the present disclosure,
quantifying
the user-executed movement comprises quantifying a duration, speed and angle
of the user-executed movement.
[0080] For example, according to methods of the present disclosure, the
plurality
of sensor are fixed to a stick, wherein the received measurements indicate a
motion of the stick, and wherein quantifying the user-executed movement
comprises determining a speed of an object hit by the stick.
[0081] For example, according to methods of the present disclosure,
quantifying
the user-executed movement is further based on at least one additional set of
predetermined motion metrics chosen from swing amplitude, swing duration,
stick speed, stick acceleration, wrist effect, flex analysis, motion
amplitude, and
motion behavior.
CA 02860129 2014-08-20
[0082] For example, according to methods of the present disclosure,
quantifying
the user-executed movement is further based on at least one additional set of
predetermined user metrics chosen from user age, user gender, user location,
user height, user weight, stick length, stick model and stick flex.
[0083] For example, according to methods of the present disclosure, the
plurality
of sets of predetermined metrics are motion patterns.
[0084] For example, according to methods of the present disclosure, when the
received measurements substantially matches more than one set of
predetermined metrics, identifying the type of the user-executed movement is
based on correlation of the received measurements with at least one additional
set of predetermined metrics.
[0085] For example, according to methods of the present disclosure,
characterizing the user-executed movement comprises detecting a motion
starting event when a first subset of the received measurements substantially
corresponds to one of a plurality of sets of predetermined starting event
metrics,
each set of predetermined starting event metrics being associated with one or
more event continuation metrics; determining a presence of a motion
continuation event when a second subset of the received measurements
received after the first subset substantially corresponds to one of the one or
more
event continuation metrics associated to said one of the plurality of sets of
predetermined starting event metrics; said one of the event continuation
metrics
being associated with one or more event completion metrics; and determining
the
presence of a motion completion event when a third subset of the received
measurements received after the second subset substantially corresponds to one
of the one or more event completion metrics associated to said one of the
event
continuation metrics.
[0086] For example, according to methods of the present disclosure, the motion
starting event is a backswing of a stick, the motion continuation event is a
downswing of a stick and the motion completion event is an impact of the stick
with an object.
CA 02860129 2014-08-20
16
[0087] For example, according to methods of the present disclosure, detecting
the motion starting event comprises detecting from the received measurements a
motion pattern associated with the backswing of the stick; determining the
presence of the motion continuation event comprises monitoring received
acceleration measurements and detecting that the received acceleration
measurements exceeds a predetermined acceleration threshold; and determining
the presence of a motion completion event comprises monitoring over an impact
detection period received measurements and detecting that the received one or
more measurements exceeds an impact threshold.
[0088] For example, according to methods of the present disclosure, the
plurality
of sensors are operable to measure acceleration in three axes in a range of
approximately OG to 16G.
[0089] Many human movements associated with sport are particularly
complex, especially when a player use a sport instrument such as a stick or a
racket for contact on a ball or other objects. In hockey, a slap shot combines
fine
and precise movements at high accelerations, especially as the hockey stick
hits
the playing surface before hitting the puck. During a slap shot, the two upper
members are in mechanical connection with a first object, the stick, to
interact
with a second object, the puck and surface made of ice, hard floor or turf.
Accurately capturing the entire movement sequence involved in a slap shot
therefore requires measurement of both low accelerations (less than 3G)
movements at a great precision and measurement of high accelerations (greater
than or equal to 3G, such as tens of G) requiring less accuracy. Such
characteristics of movement can be found, with necessary adaptations, in other
sports and movements such as golf swing, baseball batting and tennis, among
other.
[0090] Currently, no single inertial sensor or accelerometer on the market
offers a wide dynamic range enabling the concurrent capture of high and low
acceleration movements at high resolution. Such an inertial sensor would have
CA 02860129 2014-08-20
17
technical specifications able to provide an optimized balance between accuracy
of motion detection and wide dynamic range.
[0091] With respect to a sport accessory designed to hit another object,
such as a hockey stick, the ideal positioning for an inertial sensor, in terms
of
motion capture, would be close to the point where the performance metrics are
easier to measure. For an hockey stick, the lower extremity of the stick as
close
to the hockey stick blade as possible, where acceleration and movement
amplitude are maximal remains the intuitive placement. However, positioning a
sensor on or near a hockey stick blade is impractical due to very frequent
mechanical shocks affecting the stick that would damage, temporarily or
permanently, the sensor. Most commercially available inertial sensor units are
temporarily disabled by strong impact, where a sharp "shake" movement induces
unreliable measurements. Implementing a sensor inside the stick near or on the
hockey stick blade would require structural modification to the stick that
would
prevent user friendly retrofitting in existing sticks. Any weight addition
located far
from the hand placement tends to unbalance the stick in a way appreciable to
most hockey players.
[0092] Consequently, no prior art system provides a motion tracking
device that can be easily and practically integrated in the handle section of
a
sport instrument such as a hockey stick, and therefore protecting the inertial
sensor against mechanical shocks, without compromising in movement tracking
performance. Such a device would be simple of operation and user friendly,
such
that a player could easily and rapidly install the device in any stick having
a
hollow section.
[0093] As provided by various exemplary embodiments disclosed herein,
installing the motion sensor inside the handle of existing sticks such as
hockey
sticks would maximize user friendliness while insulating the sensors from most
mechanical shocks. A majority of hockey sticks are made of composite material
with a hollow section where a sensor can be installed. However, positioning
motion sensor near, on, or inside the handle part of the stick is counter-
intuitive in
CA 02860129 2014-08-20
18
terms of motion capture performance. The handle follows the angular motion of
a stick, at a fraction of the acceleration felt by the opposite end of the
stick.
Moreover, hockey sticks are made of flexible material enabling the stick to
gather
important tension when hitting the ice, before liberating tension as kinetic
energy
when hitting the puck, for maximal puck speed and movement efficiency. It may
be impractical to quantify the dynamic behaviour of the stick by only
considering
acceleration on the handle part of the stick, where the stick does not suffer
bending.
[0094] After measurement, acceleration data can be processed by
algorithms to first identify and then quantify movements, and acceleration
data
can be adapted so as to take optimal advantage of the accelerometer
capabilities.
[0095] It is a challenge to offer a generic movement tracking devices
intended for use by any player of a given sport, irrespective of their level
of
practice in such sport, while providing a consistent accuracy of movement
quantification and easiness of use. Other technical difficulties lay in the
nature of
performance metrics to be presented to different types of users of a given
sport.
The value of measurement constants specific to each user (per example, length
of arms or legs) must be taken in consideration. The range of motion variation
for a typical movement performed across a large population of users must be
considered. Yet other technical difficulty lay in the format under which the
captured movement data can be presented to users in order to provide useful
and accurate quantification of movement representative of the user's
performance.
[0096] A need of the market that is not fully addressed by prior art is to
provide a motion tracking system based on inertial sensors offering a wide
dynamic range enabling the concurrent capture of high and low acceleration
movements at high resolution that is optimized for complex gesture tracking.
[0097] A need of the market that is not fully addressed by prior art is to
provide a motion tracking system based on inertial sensors which can detect,
CA 02860129 2014-08-20
19
discriminate, identify and quantify gestures automatically among a plurality
of
heterogeneous non-choreographed movements.
[0098] Another need of the market that is not fully addressed by prior art
is
to provide a portable motion sensor which can automatically operate by
recognizing movement, without the need for prior manual activation, that does
not require any intervention from the player for its operation, and that can
be
used seamlessly over a long period of time.
[0099] Another further need of the market that is not fully addressed by
prior art is to provide an inertial sensor unit of solid construction that can
withstand the mechanical stress imposed on a sport instrument such as a hockey
stick, that is easy to install, to calibrate and to operate with a minimum of
operations while providing a high accuracy of movement quantification.
[00100] The systems, devices and methods for quantifying motion
described herein according to various exemplary embodiments at least partially
address these shortcomings.The systems, devices and methods for quantifying
motion described herein according to various exemplary embodiments are
intended to meet the needs of sport practitioners to better understand their
performance during practice or competition, through movement tracking,
movement analysis and quantification of movements, especially non-
choreographed movements such as those encountered in racket sports and team
sports.
[00101] The systems, devices and methods for quantifying motion
described herein according to various exemplary embodiments may be applied
for sports involving the use of a stick, a bat or a racket for contact with
another
object, such as a ball, a puck or another object of the same type, such as per
example and without limitation ice hockey, field hockey, baseball, cricket,
polo,
golf, tennis, badminton and lacrosse.
[00102] According to various exemplary embodiments, a device for motion
identification includes an enclosure and a plurality of sensors being provided
with
the enclosure and configured to measure acceleration in three axes in a first
CA 02860129 2014-08-20
acceleration range, acceleration in three axes in a second acceleration range,
and angular motion in three axes.
[00103] Referring now to FIGS. 1-2, a motion sensor unit 10 comprises an
inertial sensor 20, a microcontroller 30 which executes program instructions
to
detect, interpret and quantify motion data, a memory means 40 to store data
including motion data and performance data and a telecommunication means 50,
a clock 60, a battery 70, a board 80 that holds the electronic components
mentioned above and a protective enclosure 90.
[00104] The microcontroller 30 described herein may be implemented in
hardware or software, or a combination of both. It may be implemented on a
programmable processing device, such as a microprocessor or microcontroller,
Central Processing Unit (CPU), Digital Signal Processor (DSP), Field
Programmable Gate Array (FPGA), general purpose processor, and the like. In
some embodiments, the programmable processing device can be coupled to
program memory, which stores instructions used to program the programmable
processing device to execute the controller. The program memory can include
non-transitory storage media, both volatile and non-volatile, including but
not
limited to, random access memory (RAM), dynamic random access memory
(DRAM), static random access memory (SRAM), read-only memory (ROM),
programmable read-only memory (PROM), erasable programmable read-only
memory (EPROM), electrically erasable programmable read-only memory
(EEPROM), flash memory, magnetic media, and optical media.
[00105] The inertial sensor 20 is a sensor arrangement comprising a
plurality of motion sensors, typically an accelerometer group 22 for measuring
acceleration in three axis, a gyroscope group 24 for measuring angular motion
in
three axis and optionally magnetometers 28 for measuring the strengths and
direction of magnetic fields. The accelerometer group 22 comprise a plurality
of
accelerometers, for example in groups of three accelerometers for multi axis
measurements, of different ranges: high acceleration (ex: greater than or
equal to
about 10G, greater than or equal to about 15G, about 10G to about 20G,
CA 02860129 2014-08-20
21
between about 10G and about 20G, or tens of G for measuring high impacts),
small acceleration (ex: less than about 3G, less than about 10G, or less than
about 15G for high accuracy measurement of motions), high frame rate (higher
than 1000 measurements per second for high resolution motion analysis), low
frame rate (less than 1000 measurements per second, for low frequency motion
detection and reduced power consumption), etc. For example, one group of three
accelerometers may be operable to measure acceleration in three axes in the
high acceleration range and in the small acceleration range. For example, the
one group of three accelerometers may be operable to measure acceleration in
three axes in an acceleration range of about OG to about 16G. Technical
requirements similar to those of accelerometer group 22 apply, with the
necessary adaptation, to gyroscope group 24. The accelerometer group 22 and
the gyroscope group 24 are typically assembled as closed as possible to one
another on a single chip or on two distinct chips on board 80. Optionally,
inertial
sensor 20 can comprise one or several magnetometers 28. Since
accelerometers have limited precision in rotational movements at constant
speed, directional information may be obtained from magnetometer 28 or an
electronic compass (not shown), thereby enhancing the precision of recorded
data.
[00106] The raw sensor data provided by the inertial sensor 20 is received
by a data acquisition system integrated to microcontroller 30. The data is
then
stored in memory means 40 or transmitted to a remote computer 100 such as a
handheld device or a smart phone, or another external data reception system
accessible locally or based on a remote network, preferably via wireless
communication using telecommunications means 50.
[00107] Remote computer 100 is typically installed with an operation
system, software applications and logical device with multiple processing
cores
and/or CPU arrangement. The remote computer 100 described herein may be
implemented in computer programs executing on programmable computers,
each comprising at least one processor, a data storage system (including
volatile
and non-volatile memory and/or storage elements), at least one input device,
and
CA 02860129 2014-08-20
22
at least one output device. For example, and without limitation, the
programmable computer may be a programmable logic unit, a mainframe
computer, server, and personal computer, cloud based program or system,
laptop, personal data assistance, cellular telephone, smartphone, or tablet
device.
[00108] Each program is preferably implemented in a high level procedural
or object oriented programming and/or scripting language to communicate with a
computer system. However, the programs can be implemented in assembly or
machine language, if desired. In any case, the language may be a compiled or
interpreted language. Each such computer program is preferably stored on a
storage media or a device readable by a general or special purpose
programmable computer for configuring and operating the computer when the
storage media or device is read by the computer to perform the procedures
described herein.
[00109] Memory means 40 is may be a RAM (random-access memory),
DRAM or SRAM memory unit such as a flash memory drive. Communications
means 50 is typically a wireless radio communication modem such as Bluetooth
modem, a WiFi modem, an ANT+, a near-field communication modem, etc.
Optionally, the motion sensor 10 comprise a coprocessor 52 dedicated to wired
or wireless communication with a specific external computer, such as per
example an authentication coprocessor by company Apple, inc. that enable
connection between the sensor unit 10 and an device of the iPhone product
line.
Clock 60 is a real-time clock (RTC) electronic circuit that keeps time and
allow
time-related labelling of motion event quantifying by the microcontroller 30.
Battery 70 is used to power the inertial sensor 20 and other circuits of the
motion
sensor unit 10, and comprise a miniaturized battery, such as a Lithium Polymer
battery which capacity is defined by the required operation and standby time.
Exemplary battery capacities range from 100mAh to 500mAh and enable
operation time ranging from one to two hours to over thirty hours. A universal
connector 71 on the top face of the enclosure 90 connected to microcontroller
30
CA 02860129 2014-08-20
23
allows recharging of the battery and connection to an external computer or
hand
held device for software updates and upgrades, and data transfer.
[00110] The sensor unit 10 may also include a battery status monitor 72
comprising a circuit 73 and a LED indicator 74 that is affixed on the external
top
face of the enclosure 90. A display panel 75 and an on/off switch (not shown)
are also located on the top face of enclosure 90. Display panel 75, typically
a
liquid crystal display or OLED display is connected to microcontroller 30 and
is
used to provide user information such as memory means 40 status, and
performance data such as movement speed, number of movements in memory,
and status of a connexion to an external device using communications means 50
or the universal connector 71.
[00111] The motion sensor unit 10 is made of standard components
available on the market from manufacturer such as ST Microelectronics Inc.
(Scottsdale, AZ) or Analog Device Inc. (Norwood, MA) for the accelerometer 22,
ST Microelectronics Inc. (Scottsdale, AZ) or InvenSense (Sunnyvale, CA) for
the
gyroscope 24, Texas Instruments Inc. (Dallas, TX) or Atmel Corporation (San
Jose, CA) for the microcontroller 30 and memory means 40, and Roving
Networks, Inc. (Los Gatos, CA), Texas Instruments Inc. (Dallas, TX) or
Blueradios Inc. (Englewood, CO) for the communication means 50.
[00112] As shown in FIG 2, the enclosure 90 is adapted to fit inside the
hollow section of a stick (not shown) using stopper parts 92 located at the
bottom
and top sections of the enclosure. Each stopper part 92 comprises several
tongues 94 located on each of the four sides of the stopper part and enable a
tight surface contact with the internal sides of the stick's hollow handle.
Enclosure 90 is made of a material such as plastic or PVC (Polyvinyl chloride)
moulded plastic that is produced easily using, for example, injection moulding
techniques. Installation of the sensor unit 10 requires removing the cap that
closes the hollow section at the end of the stick, placing the sensor unit in
the
hollow section and pressing on the unit, the stopper parts 92 providing a
CA 02860129 2014-08-20
24
progressive pressure, until only the top part of enclosure 90 covers the top
part of
the stick handle.
[00113] According to various exemplary embodiments, locating the sensor
unit 10 in the handle part of a stick permits maximal user-friendliness and
ease of
installation, per example easy retrofitting in existing hockey sticks,
allowing any
user to install the sensor unit 10. Positioning the sensors on the stick
handle
also offer protection of the sensors against shocks.
[00114] According to various exemplary embodiments, the sensor unit 10
can be connected to and communicate with other, external sensors (not shown)
which data are processed as raw data by the microcontroller 30. Said external
sensors can be positioned on the stick or on the users' wear or equipment
including per example and without limitation helmet, shoes, skates various
protective equipment pieces, belt, wrist bands, thigh bands, arm bands,
glasses,
etc. Said external sensor such as a vibration sensor, an accelerometer, a
gyroscope, temperature sensor, humidity sensor, flexion sensor, orientation
sensor (such as magnetometer), a microphone, a camera, a video camera, a
proximity sensor, any combination, heterogeneous or homogeneous, of multiple
instance of sensors, etc. Such a sensor can be per example a bending sensor or
a flexion affixed on or moulded on the blade part or on the shaft part of the
stick
to recognise mechanical bend of the appropriate part.
[00115] Motion data acquisition
[00116] According to an exemplary method for motion identification,
measurements taken during an user-executed movement by the plurality of
sensor are received and the user-executed movement is characterized based on
the received measurements.
[00117] According to various exemplary embodiments, the microcontroller
30 is configured to identify and quantify several movement metrics, using the
inertial sensor 20 located on a single measurement point located in the handle
section of a stick or on another point maximizing the ergonomics of the
product
implanting the sensor unit 10. Inertial sensor 20 is configured for
positioning
CA 02860129 2014-08-20
near the handle part of a stick, without reduction in spatial sensing
performance
compared with positioning inertial sensors on the end section of the stick
such as
a hockey stick blade or a golf club head. Such arrangement raises technical
difficulties not addressed by prior art, which require specific motion
analysis and
processing. Such sensor placement, distant from the contact point where most
accurate speed measurement would be expected, requires more processing to
compensate for the indirect measurements. For example, the algorithm requires
user specific information, such as information related to the "length",
"curve" and
"flex" of the hockey stick, in order to compute accurate speeds, accelerations
and
rotation speeds.
[00118] The inertial sensor 20 may provide a more efficient execution in
real-time of detection algorithms (described below) for the analysis and
quantification of complex movements. The inertial sensor 20 may be calibrated
for the quantification of hockey movements but can be used for other sports
requiring the use of a stick in complex movements such as golf, baseball,
lacrosse, tennis, polo, badminton, etc.
[00119] The method for motion identification may be carried out wholly on
the sensor unit 10, wholly at the remote computer 100 or partitioned between
the
sensor unit 10 and remote computer 100. For example, portioning the carrying
out of the method between the sensor unit 10 and remote computer 100 may
improve the amount of local processing on the sensor unit 10, with respect to
the
required data storage capacity, quantity of data to be exchanged between the
sensor and remote computer 100, communication bandwidth and power
consumption. According to the exemplary method, characterizing the user-
executed movement includes correlating the received measurements with the
sets of predetermined metrics and detecting whether the received measurements
substantially matches one of the sets of predetermined metrics. The sets of
predetermined metrics may be stored on the memory of the sensor unit 10, at
the
remote computer 100 or both. The sets of predetermined metrics may be sets of
measurements that are representative of types movements that the user is
expected to execute. For example, each set of predetermined metrics may be
CA 02860129 2014-08-20
,
,
26
associated with a movement type. For example, the predetermined metrics may
be determined during a training stage where the different movements are
performed while being measured by a plurality of sensors as described herein.
For example, the predetermined metrics may be motion patterns.
[00120] For example, in a first "motion detection" step, the
algorithm
executed by microcontroller 30 processes motion data acquired by the inertial
sensor 20 in real time to detect for the presence of "known" motion patterns
by a
correlation process involving motion patterns (defined below) stored on memory
40. For example, new motion patterns can be "learnt" or acquired by loading a
pre-existing library of motion patterns, or by using the sensor unit 10 in
recording
mode while specific motions are executed by user.
[00121] A motion pattern includes a string of motion data along
time
coordinates. Motion data refers to the raw measurement data provided by the
accelerometer group 22 (accelerations in x, y, z), the gyroscope group 24
(angular speed variation, roll, pitch yaw) and optionally the magnetometer 28
(magnetic field values, Bx, By, Bz). Other sensors including without
limitation
spatial data, vibration, temperature, heart beat, blood pressure, or wind
speed
sensors, can be implanted in the sensor unit 10 and sensing data from these
sensor can be used as raw data. These raw data are acquired at a sampling rate
that can vary generally from several hertz (Hz) to several hundred of hertz
(Hz).
The best performing sensors can provide data over a thousand hertz (kHz)
sampling rates. The sampling rate is usually configurable for each sport or
application. Raw data is quantified on 8, 12 or 16 bits, varying with the
quality of
sensors and MEMS components.
[00122] According one exemplary embodiment, the microcontroller
30
focuses during the motion detection phase on a pre-established number of
metrics that are representative of the motion patterns associated with
movements typically performed during a given sport. A metric can be part of a
motion pattern, such as a spatial acceleration or movement along x-y or x-y-z
plans during a given time interval, or series of such accelerations or
movements.
CA 02860129 2014-08-20
27
Proper configuration allows the sensor unit to focus on the reduced number of
different metrics that are representative of a motion pattern. Processing a
lower
number of metrics reduces computation requirements at the microcontroller 30,
data storage requirement and power consumption. A fast motion will require a
limited volume of motion data and small memory 40 requirements, while longer
or slower motions require a greater amount of memory.
[00123] Motion patterns and metrics may be programmed as part of the
executable source code run by the microcontroller 30. Each metric may be
specific to a motion pattern found in a sport application.
[00124] According to various exemplary embodiments, characterizing the
user-execute movement further comprises identifying the type of the user-
executed movement. For example, this identifying may be carried only when a
match with a set of predetermined metrics has been identified. For example,
the
identifying may be carried out based on the correlation with the set of
predetermined metrics used for the detecting. Alternatively, the identifying
may
be carried out based on correlation of the received measurements with at least
one additional set of predetermined metrics. For example, each of the
additional
set of predetermined metrics may be associated with a movement type.
[00125] For example, in a second "motion identification" step, the
algorithm
executed by microcontroller 30 runs a gesture detection routine that
characterizes particular movements, in order to identify the type of movements
performed by user, by correlation, among a library of typical movements
associated with a given sport. The algorithm then confirms if a motion pattern
qualifies or not as a typical movement to be quantified. During the motion
identification step, the algorithm also discriminates and quantifies motions
among
motions that share same or equivalent motion patterns. This second step is
facultative as some motions are de facto identified by the pattern recognition
process of the first phase. Per example, taking ice hockey as an illustration,
a
"slap shot" movement can be precisely identified through pattern recognition
during the motion detection phase because of its distinctive signatures
involving
CA 02860129 2014-08-20
28
high elevation of the stick blade, high movement amplitude and fast
accelerations, as no other hockey motion shares equivalent characteristics.
However, a "snap shot" motion and a "passing" motion can share several metrics
such as duration, stick angle and acceleration, and may be considered as
similar
motion patterns. In such case a more detailed sample analysis is performed by
the algorithm, using a larger set of motion metrics that are pre-established
for
each motion pattern. Hockey movement that can be identified by the system as
motion patterns includes notably "wrist shot", "snap shot", "slap shot",
"dribble",
"assist", "passing". For some hockey movements, such as passes and assists,
the system only registers the time of movement and number of movement
occurrence thus the
algorithm increases value of the motion counter, and
return to the motion identification step. For other hockey movements, a motion
quantification must be perform in order to compute performance metrics with
meaning to the user.
[00126]
According to various exemplary embodiments, characterizing the
user-execute movement further comprises quantifying the user-executed
movement based on the received measurements.
[00127] For
example, in a third "motion quantification" step, microcontroller
30 executes a second algorithm that quantifies the motion executed by user.
The second algorithm is executed automatically when a movement is identified
as per the motion identification step, using raw data stored on memory 40
associated with the given movement. The second algorithm calculate quantified
movement data including, using a hockey shot as an illustration, the power of
a
throw, the duration of a movement, the amplitude of the angle of the momentum
gesture. In addition, the second algorithm computes an estimate of the maximum
speed of the puck. This estimate provides an approximation of the real speed
of
the puck immediately after impact by the stick.
[00128] Once the
previous three steps are completed, the microcontroller
30 stores the computed quantified movement data and the raw data associated
with movements identified and quantified, in memory 40 or transmit to the
remote
CA 02860129 2014-08-20
29
computer 100 using telecommunication means 50, in batch or for each individual
movement. In the absence of communication link such as wireless network, the
quantified movement data is stored on memory means 40, which is typically able
to store data representative of several thousand movements.
[00129] According to various exemplary embodiments, the method further
includes determining a speed of an object hit by a stick being used by a user
in
executing the movement. For example, an algorithm executed on microcontroller
30 can estimate the information on the movements of a second object that is
not
mechanically connected to the stick. The sensor unit can use estimation
algorithms to compute motion information of a second object. For example, a
sensor unit located on a hockey stick can estimate the speed achieved by the
puck once hit by the player. In a possible embodiment, the sensor unit would
compare the motion metrics from a single puck hit, such as a "slap shot", to
pre
recorded data that have been correlated with puck speed measurement (using
per example, a radar unit), in order to identify the most probable puck speed.
In
order to yield more accurate puck speed estimates, correlated data can
include,
in addition to motion and speed, player centric variables such as physical
measurement, level of play, age, sex, and other player characteristics which
typically impacts puck speed achieved with slap shots, that otherwise share
similar motion metric when performed by different players. In another possible
embodiment, a formula is used to directly compute puck speed estimation from
pre recorded and correlated data. Said formula is developed using known linear
regression techniques and/or known genetic algorithms. In order to increase
the
precision of the estimation, said linear regression or genetic algorithm can
be
computed for smaller reference groups, groups composed from boundaries or
criteria such as player age, player level, player size and weight, etc. In all
cases,
the estimations assume an accurate puck hit and correct puck trajectory. In
yet
another embodiment, stick related information is considered by the algorithm,
such as length, material, weight, blade length, blade angle and flex.
[00130] According to various exemplary embodiments, using the system in
recording mode, the sensor unit 10 records raw sensor data provided by the
CA 02860129 2014-08-20
inertial sensor 20 while the user executes specific motions with the stick.
Microcontroller 30 then executes algorithms to process the recorded sensor
data
to identify key distinctive elements of the motion patterns. The recorded data
can
alternatively be processed by an external processing unit such as remote
computer 100 which executes same algorithm, when the complexity of the
calculation exceeds microcontroller 30 processing capabilities. The step of
identifying key elements of a motion pattern is provided by a detection
process,
using key motion markers for real-time motion pattern recognition, and a
quantification process, using formulas for performance metric computation. A
key motion marker is a measurable value, which can be computed and identified
in real-time, and which corresponds to a specific posture, motion metric or
movement in a motion sequence. In the identification of slap shot motions, the
motions "start", "downswing start" and "impact", as illustrated in Figs 3A-C,
are
examples of key motion markers, identified by analysing in real time the
monitoring metric. A performance metric computation formula is an equation
taking input data such as sensors' raw data or motion metrics, like the
angular
speed of the stick and acceleration at specific instants during a slap shot
motion,
to provide meaningful outputs in a given sport, such as in ice hockey the
duration
of a complete motion, the angle in degrees of the backswing and the speed of
the hockey stick.
[00131] As an example, the recording mode can be used to calibrate or
tune the sensor unit to a user specific motion signature. It can also provide
a
means to use the sensor unit in a different sport, as per example, from hockey
to
lacrosse, where motions are different but share similarities.
[00132] Device activation
[00133] For ergonomic enhancements, or for robustness concerns, the
sensor unit 10 can integrate an automatic start/stop mechanism, where the "on"
command is triggered by a motion, e.g. an impact of at least 4G in one or any
dimension, while the "off" command can simply be triggered by a timeout, per
CA 02860129 2014-08-20
31
example a duration of 30 seconds without movement. Such parameters can also
be configured or turned off by user.
[00134] According to various exemplary embodiments, the plurality of
sensors are operable between a sleep state and a wake state, wherein at least
one of the sensors remains in a wake state while the remainder of the sensors
are operating in a sleep state, and wherein detection of a motion by the at
least
one sensor in the wake state triggers activating at least one of the remainder
of
the sensors to the wake state.
[00135] For example, at least one sensor of either the accelerometer group
22 or the gyroscope group 24 is kept powered in wake mode at all time, when
the
sensor unit 10 is otherwise turned off, wherein said one sensor can detect a
short
movement such as per example and without limitation, a "shake" motion
vertically, laterally or rotationally, which movement will trigger a powering
on of
the sensor unit 10. A shake movement is perform typically when user gets hold
of or collects a stick that was stored or in an immobile position.
Accordingly, user
does not need to notify the motion sensor unit 10 that it will launch. The
sensor
unit is "listening" to all actions of the user at all time.
[00136] According to various exemplary embodiments, the device or system
further includes an identification tag reader operable to read an
identification
code, such as a bar code or an radio-frequency identification tag (RFID tag).
[00137] As an illustration of the device activation and motion
identification
and quantification capabilities, a golf user would not need to configure the
motion
sensor ; the sensor unit 10 would know which club is used (provided a sensor
is
installed in each club or that the sensor unit is provided with a mean of
identifying
specific golf clubs like barcodes or RFID tags, such as an infrared bar code
reader or a radio transceiver, in which case the sensor unit can be affixed on
the
exterior surface of the stick for convenience, or wear by the player,
typically on a
glove or on a wrist), and exactly when user executes a drive (notably in
detecting
the vibrations and impact associated with the club impacting the ball).
CA 02860129 2014-08-20
32
[00138] Improvements provided by various exemplary embodiments
described herein may include (i) faster data acquisition by the use of an
inertial
sensor 20 comprising plurality of sensors of high and low sensitivity and high
sampling rate, including accelerometers 22, gyroscopes 24, magnetometers 28
and other sensors ; (ii) acquisition of higher quality data by the
identification of
non-choreographed movements using algorithms executed by microcontroller 30
that identifies and validates motion metrics by comparing raw data from
inertial
sensor 20 to a bank of motion metrics on memory 40 ; (iii) richness of the
motion
data captured and quantified, by the use of algorithms executed by
microcontroller 30 to quantify identified motion metrics and to provide usable
information such as movement speed, ball or puck speed estimate, duration of
movement, number of movements, etc. (iv), better ergonomic by positioning the
sensor unit in the hollow section of the stick, where the sensor unit is not
obstructing any user movement and is protected against shocks ; (v) simplicity
of
installation by the user by simply sliding the sensor unit inside the hollow
section
of the stick ; (vi) continuous operation where the sensor unit can operate at
all
time and active itself automatically without any operation other that a simple
movement such as lifting the stick ; (vii) richness of the motion
quantification
experience, by enabling the sensor unit to communicate with an external device
where quantified motion information can be presented, communicated, compared
and exchanged, notably using applications localised on hand held devices and
using social networks.
[00139] Referring now to FIGS. 3A-B-C, examples of motion patterns
associated with a hockey slap shot or snap shots are provided and exemplified
through graphs where duration or timeline is represented in the x axis and
spatial
movement or acceleration is represented on the y axis. Said graphs are
rendition of the combined raw data provided by the inertial sensor 20 and
interpreted by the algorithms executed on microcontroller 30.
[00140] As represented by FIG 3A, a shot motion is identified comprising
the following motion patterns successively: "initial position", "backswing",
"downswing", "impact with the puck", follow "through" and back to an "initial"
or
CA 02860129 2014-08-20
33
neutral position. The motion detection starts with a backswing of high
amplitude
identified as event no 1, as shown on the graph at time 6.65, followed by a
sharp
downswing movement starting in event 3 culminating in a impact with the puck
as
event no 5 and by movement follow thorough at event no 7, after which the
system is idle, until a new event is detected as illustrated by the second no
1
event at time 6.87+. For example, the method is carried out by detecting
and/or
identifying a slap shot movement by the value of motion metrics typically
characteristic of a slap shot associated with a backswing of events no 1,
followed
by a backswing delay identified as lapse no 4, a downswing start detection of
event 3, a downswing delay identified as lapse no 7, and impact of event no 5,
the full duration of the quantified movement is identified as lapse no 2 and
the
moving average is provided as event no 6. The moving average is an average
made on a sliding window of a specific number of samples. For every new
sample, the average is recalculated with the new sample, while dropping the
last
one of the window. Such moving average is used to confirm a strong trend
within
highly dynamic data stream.
[00141] As represented by FIG 3B, a hockey shot is often characterized by
an impact with ice and the puck successively, as shown by the three peaks
provided by the monitoring metric graphs, said peaks being associated with the
"ice contact", "puck contact" and "release" movements and constitutes motion
metrics identified and quantified. Using a stick with a certain flex
coefficient, the
stick will accumulated energy during the ice contact and puck contact
movement,
referred to as "stick loading" and is characterized notably by the duration,
distance and measurement period.
[00142] For example, the sensor unit can estimate a ball or puck speed
following a swing movement based on raw data provided by the inertial sensor
using motion metrics from a group including movements and values such as
"swing amplitude", "swing duration", "stick speed", "stick acceleration",
"wrist
effect", "detailed flex analysis", "motion amplitude" and "motion behaviour".
The
precision of said puck speed estimate can be increased by the specification of
user centric data using the application located on the remote computer and
CA 02860129 2014-08-20
34
transmitted to the microcontroller 30 using the telecommunications means 50,
such user centric data including without limitation, "user age", "user
gender",
"user location", "user height", "user weight", "stick length", "stick model",
"stick
flex".
[00143] As represented by FIG 30, a hockey shot is often characterized by
motion metrics such as pre impact downswing and post impact downswing,
before a return of the stick to an initial or idle position.
[00144] According to various exemplary embodiments the user-executed
movement is characterized by detecting a motion starting event when a first
subset of the received measurements substantially corresponds to one of a
plurality of sets of predetermined starting event metrics. Each set of
predetermined starting event metrics may be further associated with one or
more
event continuation metrics. Then the presence of a motion continuation event
is
determined. For example, there is presence of a motion continuation event when
a second subset of the received measurements received after the first subset
substantially corresponds to one of the one or more event continuation metrics
associated to said one of the plurality of sets of predetermined starting
event
metrics. The sets of event continuation metrics may be further associated with
one or more event completion metrics. The presence of a motion completion may
be further determined when a third subset of the received measurements
received after the second subset substantially corresponds to one of the one
or
more event completion metrics associated to the one of the event continuation
metrics.
[00145] For example, in the context of a hockey movement, the motion
starting motion starting event is a backswing of a stick, the motion
continuation
event is a downswing of a stick and the motion completion event is an impact
of
the stick with an object. Accordingly, detecting the motion starting event
includes
detecting from the received measurements a motion pattern associated with the
backswing of the stick. Determining the presence of the motion continuation
event comprises monitoring received acceleration measurements and detecting
= CA 02860129 2014-08-20
that the received acceleration measurements exceeds a predetermined
acceleration threshold associated to the downswing of the stick. Furthermore,
determining the presence of a motion completion event includes monitoring over
an impact detection period received measurements and detecting that the
received one or more measurements exceeds an impact threshold.
[00146] The method of movement quantification step according to one
example is performed by algorithms executed by the microcontroller shown
graphically in the flowcharts of FIGs. 4A-B, as is illustrated by the
detection, and
identification of a shot movement. The "A" bubble indicates a continuation
between FIG 4A and 4B.
[00147] In a first step "Activation Trigger" 210, the sensor unit
is
automatically activated by a movement or user action on the stick, such as an
acceleration or rotation.
[00148] In a second step "Shot Start Detection Process" 212,
following
activation, the sensor unit initializes metric values and starts the real-time
monitoring and analysis of metrics. The sensor unit uses as inputs raw data
from
the inertial sensor, principally acceleration values Ax, Ay, Az provided by
the
accelerometer and angular rate values Mx, My and Mz provided by the
gyroscopes, and provides a metric value expressed in m*rad/s3 as outputs. Such
value illustrate a rotational speed multiplied by a linear acceleration, hence
the
meter*radian numerator and the secondA3 denominator. The system returns to
an idle status "B" 205, when the movement stops before qualifying for a third
step
event.
[00149] A third step "LMIN1" 214, by which the system logs as a
"local
minimum #1" event referring to the start of a potential shot motion, is
triggered
when the algorithm detects and identify a motion pattern associated to a
"Backswing Motion Start".
[00150] A fourth step "LMAX1" 216, by which the system logs as a
"local
maximum #1" event referring to a potential backswing motion acceleration, is
triggered when the algorithm detects and identifies a motion pattern that is
CA 02860129 2014-08-20
36
greater in value than a "backswing low threshold" and lower in value than a
"backswing high threshold". The system loops in this fourth step for as long
as
the delay between the detection of LMIN1 and the current time is lower than
the
Backswing Delay (BD). If such delay is reached, the system returns to an idle
status 205, when the movement identifies in step 214 fails to qualify as a
"local
maximum #1" event.
[00151] A fifth step "LMIN2" 218, by which the system logs as a "local
minimum #2" event referring to a potential shot motion continuation, is
triggered
when the algorithm detects and identifies (i) a motion pattern associated to a
potential "Downswing Motion Start" or to (ii) a motion associated with a
potential
"Backswing motion start" in which case the motion replaces the previously
logged
"local minimum #1". The system loops in this fourth step for as long as the
delay
between the detection of LMIN1 and the current time is lower than the
Backswing
Delay (BD). If such delay is reached, the system returns to an idle status
205,
when the movement identifies in step 216 fails to qualify as a "local minimum
#2"
event.
[00152] In a sixth step "Impact Detection Metric" 220, by which the system
monitors the variation of the Monitoring Metric over an impact detection
period,
and log the event as a potential impact when the variation reaches a specific
level, i.e. the "Impact Threshold" a "Downswing Motion Start". The impact must
follow a Downswing motion within a specified time period, i.e. the "Downswing
Delay" or "DD" to be considered for the next step. If the impact is not
registered
during the specified Downswing Delay, the system returns to idle mode at step
205.
[00153] In a seventh step "Shot Detection" 222, the first algorithm
confirms
an impact event when the Moving Average of the Monitoring Metric reaches a
specific level, i.e. the Shot Threshold, and the second algorithms quantifies
the
motion metrics associated with the event and computes shot statistics from the
Low Minimum #2 event to the impact.
= CA 02860129 2014-08-20
37
[00154] Various exemplary embodiments described herein rely on the
concept of "software sensor", or "Software Defined Sensor" (SDS). The SDS can
be reprogrammed to enable the identification and quantification of wide range
of
movements and gestures. As SDS is designed to connect to a computer or a
smart phone via a wireless Bluetooth link, it can be reprogrammed remotely by
a
simple update of the client application. For example, a SDS based firmware
implemented on a mobile application located on a handheld device can allow the
identification and quantification of "slap shots" and "short shots", in
relation to
hockey. Update or upgrade of the mobile application can be downloaded from a
central computer using internet or a communication network, such newer version
of the mobile application may include a SDS allowing, for example,
identification
and quantification of "wrist shots" and of the amplitude level of the end of
the
movement (to estimate the speed reached by the puck after impact). The SDS
platform for data acquisition can be adapted easily allowing the algorithms to
identify and quantify motion metrics associated with any sport. The sensor
unit
is typically used in recording mode to record the raw motion data associated
with every pertinent motion of such new sport. The collected raw data is then
processed to identify the key features of targeted motions. The result is a
new
detection process and a new quantification process that can be uploaded in the
SDS. The SDS can therefrom detect and quantify new motions.
[00155] Referring now to FIG. 5, sensor unit 10 is connected to a
remote
computer 100 such as a smart phone with a processor, memory, database and
associated software for configuring the sensor unit, storing data and for
organizing, presenting, communicating, comparing and exchanging quantified
motion information and user information representative of its profile and
performance.
[00156] According to one exemplary embodiment, the sensor unit 10
synchronizes with a remote computer 100 such as a smart phone for updating
data and usage statistics. The software application running on the smart phone
synchronises to a localised database on a memory means connected to the
smart phone, or to a remote database on an external computer 104 using a wired
CA 02860129 2014-08-20
=
38
or wireless connection of the smart phone (cellular, WiFi, Bluetooth). This
synchronization will allow the user to keep a longer history of statistics,
but also
to access performance data and other quantified motion data of members of a
wide user community. A player performance data can be then compared to
external similar data representative of other players' performance. Quantified
motion data can be then presented via ergonomic, intuitive and engaging
interfaces.
[00157] According to one exemplary embodiment, the remote computer
100
connects to an online platform 106 such as a social network that is accessible
from a smart phone or an external computer, for the purpose of communicating,
comparing and exchanging quantified motion information and user information
representative of its profile and performance. The sensor unit 10 device can
also
be connected to interactive applications on external computers 104 or hand
held
device 105 such as smart phones, that allow users, trainers and coaches to
analyze, present, share and compare quantified motion information and player
performance data.
[00158] Referring now to FIGs. 6A-B-C-D, a hand held device 102
comprising the processor(s), memory, communication means, display and
graphical interfaces necessary to execute mobile software applications for the
presentation of user profiles, performance statistics, communication and
comparison of quantified motion data and user information for a single users
and
several users.
[00159] FIG 6A provides an example of motion analysis interface
presenting ice hockey "shots" motion events metric, for each distinct motion
metric, such as a "Slapshot", as is shown in the central window of the hand
held
device interface. The motion analysis interface provides quantified motion
data
with respect to a shot motion event, such as "shot speed", "duration",
"angle",
estimation of "puck speed", "acceleration" and "rotation speed".
[00160] Interfaces example shown in Fig 6A and 6B provides common
features such as the possibility of viewing in "Live" mode, "Shots" mode or
CA 02860129 2014-08-20
39
"Events" mode. In Live mode, quantified motion metrics and events metrics are
presented in real time, as illustrated in Fig 6A. In Shots mode, quantified
motion
metrics and events metrics are represented by lists of "shots" events of a
similar
category, as classified according to one criteria among a group of criteria
comprising "date", "speed", "duration", "puck speed", "stick speed",
"translational
acceleration" and "rotation speed". In Event mode, quantified motion metrics
and
events metrics are represented by lists of "shots events of all categories
presented together, and classified according to one criteria among a group of
criteria comprising "date", "speed", "duration", "puck speed", and
exemplifying in
the top windows, the latest or best performing events of a one or two
categories
of events such as "SnapShot/Pass" and "Slapshot" categories, where the top
performing event and average performance are represented, as is illustrated by
the interface example of Fig 6B.
[00161] User information and user performance information can be
presented using an interface on the dedicated mobile application, as is
illustrated
by the interface example of Fig 6C. Said user related information comprises
without limitation, information such as name, address, team, position and
player's
number, number of events in the system, number of followers from social medias
and number of other users follow by the user. Said user performance
information
includes "Recent events" list of events including events statistics such as
speed ;
for each event categories such as "Snapshot/pass" and "Slapshot", the total
events (shots) and game performance (number of "goals" or "assist" over number
of shot events), the top performing events defined per example by the top
speed
or another criteria, along with the average performance of all user events of
same category.
[00162] Other users of the mobile application have the capabilities to
"Follow" a user, to "Like" an event or a "Comment" an event or posting. Any
user
information or performance information can be shared on an online platform or
social media such as Facebook, using a user Facebook account linked to the
mobile application on user's remote computer or hand held device connected to
a sensor unit 10.
CA 02860129 2014-08-20
[00163] The mobile application provides a "Leaderboard" listing comparing
user's performance (virtual sport card) using user's events gathered during
game
or practice, as illustrated by the example of interface of Fig 6D. The
Leaderboard
interface provides a listing of users performance information based on at
least
one criteria such as "number" of event or "top speed", for a given event
category
such as "Slapshots" or "Snapshots/pass" category, during a given period such
as
"Last 30 days" period.
[00164] According to one exemplary embodiment, an inertial sensor unit
and a mobile application connected to an online platform such as a social
network, enable users to monitor their sport performance by identifying and
quantifying performance events, which event information and statistics can be
presented, shared, compared and commented using social medias. Per example,
each "slapshot" made by a ice hockey user will receive a score based on speed;
each user will receive a score based on the quality and power of its
"Slapshot"
score launch; this score will rank players from around the world and stimulate
healthy competition .Various exemplary embodiments described herein
mayadvantageously provide a motion tracking and quantification system, device
and method optimized for complex gestures in a non choreographed motion
sequence.
[00165] Various exemplary embodiments described herein may
advantageously provide motion tracking system based on inertial sensor
offering
a wide dynamic range enabling the concurrent capture of high and low
acceleration movements at high resolution that is optimized for complex
gesture
tracking.
[00166] Various exemplary embodiments described herein may
advantageously provide a motion tracking system which can detect, discriminate
and quantify complex gestures automatically among a plurality of heterogeneous
non-choreographed movements.
[00167] Various exemplary embodiments described herein may
advantageously a motion tracking system which can operate without the need for
CA 02860129 2014-08-20
41
a specific or prior instruction or action by the user, that recognizes
movements
automatically and which can be used seamlessly and continually over a long
period of time without the need for external systems.
[00168] Further, Various exemplary embodiments described herein may
advantageously provide a motion tracking system of solid construction that can
withstand the mechanical stress imposed on a sport instrument such as a hockey
stick, while not affecting the behaviour of the sport instrument nor hindering
user's movement in any way, while being easy to install, to calibrate and to
operate with a minimum of operations and while providing a high accuracy of
movement quantification.
[00169] Improvements provided by various exemplary embodiments
described herein may include (i) faster data acquisition by the use of an
inertial
sensor comprising plurality of sensors of high and low sensitivity and high
sampling rate, including accelerometers, gyroscopes, magnetometers and other
sensors ; (ii) acquisition of higher quality data by the identification of non-
choreographed movements using algorithms executed by microcontroller that
identifies and validates motion metrics by comparing raw data from inertial
sensor to a bank of motion metrics on memory ; (iii) richness of the motion
data
captured and quantified, by the use of algorithms executed by microcontroller
to
quantify identified motion metrics and to provide usable information such as
movement speed, ball or puck speed estimate, duration of movement, number of
movements, etc. (iv), better ergonomic by positioning the sensor unit in the
hollow section of the stick, where the sensor unit is not obstructing any user
movement and is protected against shocks ; (v) simplicity of installation by
the
user by simply sliding the sensor unit inside the hollow section of the stick
; (vi)
continuous operation where the sensor unit can operate at all time and active
itself automatically without any operation other that a simple movement such
as
lifting the stick ; (vii) richness of the motion quantification experience, by
enabling
the sensor unit to communicate with an external device where quantified motion
information can be presented, communicated, compared and exchanged, notably
using applications localised on hand held devices and using social networks.
The
CA 02860129 2014-08-20
42
person skilled in the art would understand that the various properties or
features
presented in a given embodiment can be added and/or used, when applicable, to
any other embodiment covered by the general scope of the present disclosure.
[00170] The
present disclosure has been described with regard to specific
examples. The description was intended to help the understanding of the
disclosure, rather than to limit its scope. It will be apparent to one skilled
in the art
that various modifications can be made to the disclosure without departing
from
the scope of the disclosure as described herein, and such modifications are
intended to be covered by the present document.
