Patent Grant
8986129
1. Field of the Invention
The invention relates to golfing. In particular, the invention relates to a device for analyzing and developing skills of a player. The invention also concerns a method for analyzing the performance of the player during training or playing.
2. Description of Related Art
Golf is based on one's ability to predict the trajectory of a ball in response to a certain combination of swing and club. Swings of golfers are traditionally analyzed either by naked eye during the hit or by recording the swings using video equipment and analyzing them afterwards. In both cases, for a proper analysis of the pros and cons of the swing, a professional golf trainer is usually needed. Naked-eye and video-based evaluation of swings suits well for discerning major faults in, for example, the stance and alignment of the golfer and the movement of the body of the golfer during the swing. The results of such evaluation are also highly subjective and reflect the view of the person watching the swing, typically a golf trainer. For a golf trainee, it may be frustrating if several professional have different, probably opposing, views of the pros and cons of his or her shots. What is difficult to exactly evaluate by visual observation of the golfer, is the temporal progression of the swing and especially the similarity of two or more swings. In addition, hiring of a personal trainer having a suitable video equipment is very expensive, and the video equipment is difficult to carry.
U.S. Pat. No. 6,648,769 discloses a swing analyzing system comprising an instrumented golf club including a plurality of sensors, an internal power supply, an angular rate sensor and an internal ring buffer memory for capturing data relating to a golf swing. The swings are analyzed one-by-one for assisting a golfer's swing, or for designing an appropriate golf club for a specific type of golfer.
U.S. Pat. No. 6,073,086 discloses a device for measuring the speed of swing of a movable object, such as a baseball bat or golf club. Part of the device is embedded, secured, or attached to the projectile or movable object of interest, and consists of an acceleration sensor, thresh-old circuit, and a radio transmitter.
In U.S. Pat. No. 5,688,183 a velocity monitoring system for golf clubs is described. The monitor is preferably detachably securable to the golfer's hands or a golf glove. An inferential determination of club head velocity may be made by using an accelerometer disposed in the monitor.
U.S. Pat. No. 5,233,544 describes a swing analyzing device comprising swing practice equipment such as a golf club having acceleration sensors attached all over the club for analyzing the movements of the club with good precision during a swing.
U.S. Pat. No. 4,991,850 discloses a golf swing evaluation system including a golf club containing a sensor and an associated display for indicating the force and location of the impact of the club head against a golf ball.
The prior art devices make mostly use of acceleration sensors disposed in the golf club. However, such solutions usually change the properties of the club, whereby the measurement results may be unreliable. Moreover, in the above-mentioned devices, the analysis is based on monitoring swings one-by-one. They are well suited for practicing or improving technicalities of the swing. In monitoring the swing as a whole, or increasing or analyzing the attained muscle memory needed in golf, the present solutions are more unfavorable. Golf swing is a pendulum motion, not just hitting a ball with a club. A good swing is a combination of many factors, such as a right stance of the player, good balance throughout the swing, correct spatial course of the club and feasible temporal development of the swing. In order to reduce the number of shots required, the swing of the golfer has to be very constant and regular. If one or some of the preceding factors, for example, are irregular, the whole swing, and ultimately the flight path of the ball becomes unpredictable.
There are no known devices or methods which can be used for automatically and objectively analyzing the swing of a golfer in order to assist the golfer to improve his or her skills as a golf player.
It is an aim of the invention to achieve a novel device for helping sportsmen improve their skills by providing information on the repeatability of a motor act, especially a two-phase act such as a golf swing, that is frequently needed in the sport they go in for.
In particular, it is an aim of the invention to achieve a novel device and method for helping a golfer improve his or her skills by providing information on the repeatability of the swing of the golfer.
It is also an aim of the invention to achieve a novel method for determining the repeatability of a swing of a golfer using data collected during various swings.
It is a further aim of the invention to achieve a device that assists a golfer to focus his or her training on a specific part of the swing.
It is also an aim of the invention to provide a device and method, which can be used for obtaining objective information on the courses of swings.
The device according to the invention comprises means for collecting data on courses of repeatedly performed motor acts in response to body movements of the sportsman. In addition, the device comprises means for calculating at least one characteristic number based on the data collected during the motor acts. The means for collecting data comprise at least one sensor capable of delivering information (output signal) on the body movements of the sportsman during the motor acts and a signal processing unit for refining the information delivered by the sensor. The characteristic number represents, for example, repeatability of the different parts of the swing of the sportsman, i.e., how similar the various repetitions of the motor act are to each other in terms of the parameters measured.
The method according to the invention comprises monitoring the performance of a sportsman by collecting data on the course of a plurality of motor acts performed by the sportsman. Body movements of the sportsman are is sensed by an applicable sensor. The information provided by the sensor is then refined and used for calculating at least one characteristic number representing the repeatability of the motor act.
The motor act can be, for example, a swing performed by a golfer.
The data extracted from the sensor output signal can, for example, be used to describe one, some or all of the following swing properties: the tempo of the swing, the rhythm of the swing, the duration of the backswing and the velocity of the blade of the club during downswing. In this context, the tempo of the swing stands for the total duration of the swing, i.e., the time period from the beginning of the backswing to the completion of the swing. By the rhythm of the swing is meant the temporal proportions of the back- and downswing of the whole swing. The duration of the backswing is the time period from the beginning of the swing to the turning of the swing in the upper position of the club. In estimating the velocity of the club blade during downswing, pre-stored data on the length of the club can be used. When calculating the characteristic number(s), the mean values and standard deviations, for example, of the parameters measured can be used.
After several swings, the characteristic number(s) can be calculated from the measured parameters of each of the swings. Typically this step comprises calculating the coefficients of variation, and/or other applicable statistical quantities, of the swing properties for obtaining an objective criteria on the variability of the properties, and thus the repeatability of the swing parts the parameters represent. The step can also comprise calculating a weighed sum of some or all of the coefficients of variation for obtaining a characteristic number, which represents the overall repeatability of the swing in terms of several properties at once. In this document, this kind of weighed-sum characteristic number is also called a swing index number (SIN).
Considerable advantages are achieved by the present invention. A basic challenge in many sports, especially golfing, is that the temporal and spatial course of swing varies a lot from hit to hit. This causes the accuracy of the hits to decrease. In other words, the accuracy of the hits is highly dependent on the reproducibility of the swings. We have found that a reliable analysis of swings can be carried out automatically by a suitable electronic device, which can be mounted on the body or club of the sportsman. The variability of the swings can be detected by measuring certain parameters during the swings. The parameters can be used for pointing out the potential weaknesses and faults of a golfer by calculating various characteristic numbers, which represent different sectors of the swing or the swing as a whole.
In particular, we have found that determination of the key points of time of a plurality of swings is a good tool for identifying the weaknesses of the swing. This is because for example an unbalanced stance or irregular movement of the body of the golfer reflects in temporal variations. We have found also, that at least one acceleration sensor can be used for detecting the necessary key points reliably in order to make further analysis of the swing.
Advancing golfers change their way of swinging, the grip on the club or the set of clubs every now and then. These changes often involve also some changes in the temporal course of the swing. By using the device according to an embodiment of the present invention, golfers can track these temporal changes to see which parts of the swing are getting better (more constant) and which parts should still be improved. For example, the tempo of the swing can be well maintained constant in a new swing the golfer has practiced, but the duration of the backswing can deviate more than in his old swing.
The device according to an embodiment of the invention can be manufactured light, for example, to be carried on a wrist of the player. The player may start and stop monitoring of his or her swings whenever he or she wants. The device can be used during training, for example, on a driving of chipping range, but also during playing, for example, to monitor the similarity of several drives on teeing grounds.
The determined characteristic numbers of the swing of a player have been found to correlate well with the handicap (hcp) of the player. However, the index number describes the swing of the player more exactly, because the hcp is calculated by the overall performance of a player, including also the “swingless” areas of the game, such as mental and environment-dependent aspects of the game and putting. Hence, the characteristic number, especially the swing index number, determined by means of the invention, is a reliable measure of the hitting skills of a golfer.
According to one embodiment, the device informs the player which sector or sectors of the swing should be improved in order to make the swing more repeatable. This can happen by, after a predetermined number of swings, displaying several characteristic numbers representing different properties of the swing, to the user. This enables the player to concentrate his training on the weakest sector of the swing in order to enhance his or her skills towards more accurate shots. Alternatively or additionally, a swing index number representing the reproducibility of the swing as a whole can be displayed. In addition to displaying the results, there may be implemented also some advanced features in the device. The device can, for example, detect that the duration of the backswing fluctuates unacceptably much with respect to the fluctuations in the duration of the downswing and advice the player to concentrate on clean backswings in further training, or vice versa.
As is appreciated by a person skilled in the art, the device and method disclosed in this document can be used also for evaluating the shots and serves in other sports, such as tennis and baseball. In particular, the present solutions can be utilized in all kinds of sports making use of clubs, bats, sticks, racquets or mallets. In addition, the principles of the invention are applicable, for example, in ball games that require good body coordination and reproducibility of certain moves, such as basketball and boxing. For example in basketball, the reproducibility of free throws can be analyzed, and in boxing, the characteristics of the stretches and hooks can be analyzed.
In this description, golf terms such as backswing and downswing are used for describing the phases of the movement (motor act) in question of the sportsman. A person skilled in the art easily finds equivalent phases in many other sports. For example, when serves and hits of tennis are concerned, the backswing and downswing are easily distinguishable. In basketball, the equivalent phases are the small backpull of the player with a ball in his hands in front of his face and the stretching of hands when launching the ball towards the basket. As is the case in golf, also in basketball, the success of the throw is dependent on the ability of producing an exactly identical series of motion.
In many cases, the term “backswing” can thus be replaced with one of the terms “first phase of motor act”, “preparatory step of effort” or “the step of collecting potential for a forthcoming effort”. Respectively, the term “downswing” can usually be replaced with a term “second phase of motor act”, “step of effort” or “the step of releasing the collected potential for performing an effort”.
As is evident from the preceding disclosure and the description and claims hereafter, the term “swing” is used both in the meaning of a single swing and in a broader meaning describing the general hitting performance of a golfer (as in “repeatability of swing”).
The term “course of swing/club” is used for describing the temporal course of the swing/club, including all time-related measurable quantities, such as acceleration, velocity and the spatial information on the device, on the golfer, or on the club in time. By the term “repeatability” or “reproducibility”, we mean the similarity of at least two swings as regards to the courses of the swings.
Next, the invention will be examined more closely with the aid of a detailed description and with reference to the attached drawings.
a to 1d depict four different phases of golf swing as a series of images,
In determining the preferred properties of a swing (tempo, rhythm, duration of backswing, velocity of club blade), measuring of three different time points (also called “key points”) is needed. Referring to
In
The numerals in the
21. Start the monitoring program
22. Detect the key points of a swing store predefined swing data
23. Check if a predetermined number of swings has been performed: No: go back to step 22, Yes: continue to step 24
24. Calculate coefficients of variation of predetermined properties of the swing
25. Calculate a weighed sum of the coefficients of variation calculated in step 14
26. Display results
27. End process
Starting of the monitoring program 21 is preferably done in response to user input. After starting, the device begins observing the movements of the player. In a preferred embodiment, the device includes a G-sensor or a set of G-sensors, which continuously provide data on the spatial movements of the device.
Swing detection in step 22 is preferably done using the acceleration data provided by the G-sensor(s). The device preferably detects and records the absolute or relative time points of the starting of the backswing (tα,1), the reversal of the swing in the upper position of the club (tα,2) and either the hitting of the ball or the ending of the follow-through after the downswing (tα,3). In this document, these time points are denoted tα,b,, where α is the ordinal number of the swing and b is the ordinal number of the time point. The desired properties of the swing, for example, the tempo (pα,1) and rhythm (pα,2) of the swing, the duration of the backswing (pα,3) and the velocity of the club head (pα,4), can be calculated using the recorded time points in this step or in a later step (for example, step 24). In this document, the properties determined are denoted pα,c,, where α is the ordinal number of the swing and c is the ordinal number of the property. In this example, the number of properties (C) is 4, but it can also be more or less than that. The data can be temporarily stored in a built-in or portable memory of the device. Some embodiments concerning the practical implementation of the detection step are described more closely later in this document with reference to
In step 23, the device decides, whether it has successfully detected and recorded a predefined number of swings (A) in order to calculate the characteristic number(s). The predefined number of swings is preferably 2-100, typically 5-20. The number may be given by the user or it may be preset into the memory of the device. In principle, the higher number of swings, the more statistically reliable the results are. In practice, however, the tiring and enervation of the player after a large number of similar shots may cause additional fluctuations to the measured swing parameters, whereupon the optimal number of swings can actually be quite low. It has been found, that by using the embodiments described in this document, even after a set of 10 swings, there is usually enough data to evaluate the hitting skills of a player with good accuracy.
In step 24, the mean values μc, and standard deviations σc are determined from the data set pα,c. From the means and deviations, the coefficients of variation CVc of the selected swing properties are calculated to form characteristic numbers Sc describing the reproducibility of the sub-parts c of the swing.
In an optional step 25, at least some of the quantities calculated in step 24 are weighed with weighing factors wc and summed to form a swing index number S. Thus, the swing index number S can be calculated as:
The weighing factors wc can be chosen to take into account the importance of each of the sub-parts c in formation a successful swing. The weighing factors can also be chosen such that S represents the estimated hcp of the player. It is also possible to use a special weighing function(s) fc(Sc) and/or f(S) in order to obtain results better corresponding to the real hcp of the player.
In step 26, the results of the monitoring program are displayed to the user on the display of the device. The display can comprise an LCD or TFT unit, for example. The characteristic number(s), i.e., the coefficients of variation CVc and/or the swing index number S can be displayed as pure numbers, for example, in percentages, and/or in a refined form, for example, graphically. In a further embodiment, the swing index number S in given hcp-units. Thus, the results can be displayed in the following way, for example:
In
Because coefficients of variation are calculated using statistical deviations, they are comparable despite the number of swings performed or the club used. The same applies to swing index numbers calculated from the coefficients.
In a further embodiment the user is also provided with absolute values of the properties determined. For example, the absolute the velocity of the club blade can be an interesting quantity when practicing long drives. Information can be provided on each swing individually or averaged over several swings.
Referring to
The sensor 46 comprises preferably an acceleration sensor (also “G-sensor”) or a plurality of acceleration sensors, which provide electrical signal proportional to the acceleration of the device. Such sensors can be manufactured as separate microchips, as a part of another component, or embedded in the wiring board or casing of the device. The G-sensor used can be responsive to acceleration in all spatial directions (i.e. a 3D G-sensor) or there can be arranged several (typically three) sensors sensitive to different, for example, orthogonal directions of movement. Different types of G-sensors that can be used are, for example, those based on capacitive coupling and piezoelectric effect.
The sensor or sensors 46 can me arranged to provide a plurality of signals for each direction separately (vector acceleration) or a total acceleration signal (scalar sum). In order to provide accurate information on the movements of the club, the sensor is preferably mounted on an essentially fixed location with respect to at least one part of the club. In a preferred embodiment, the sensor is an integral part of the main device, for example, a wrist watch or wristop computer, whereby the sensors follow the motion of the club precisely enough to provide sufficient data on the swing. However, we do not exclude such embodiments, which utilize at least one external sensor installed, for example, on the blade of the club and communicating with the main device by wire or wirelessly.
In addition to or instead of a G-sensor, a sensor or a set of sensors of some other type can be used. For example, position, velocity, alignment or proximity sensors can be used to provide additional data for determining, or information that is sufficient for determining, the key points of the swing. This may, however, require positioning some sensor elements separately from the main device, for example, on hitting ground, ball or a body of the golfer.
The detection of the key points of the swing is performed using the output signal(s) of the sensor(s) 46. The signals can be in analogue or digital form. The detection process can also be implemented by analogue or digital means. If digital signal processing is utilized, analogue signals can be A/D-converted by suitable electronics before further analysis. The microprocessor 41 of the device is preferably used for digital signal processing. The acceleration data can also be calibrated, filtered and/or scaled before further analysis. If several sensors or sensor channels are used, arithmetic or algebraic operations can be carried out for forming derived data, such as sum signals or vector projections, e.g. for finding radial and tangential components, of acceleration data.
The detection can made in real time as the swing proceeds, in delayed real time, or after the swing has ended by analyzing stored signal data. In a preferred embodiment, the detection is carried out such, that when a key point of a swing is detected, its point of time is “stamped”. That is, the point of time is stored in the memory of the device as an absolute value or as a relative value with respect to some other time point. In a preferred embodiment, the detection is primarily based on monitoring the value, first derivative and statistical variation of the total acceleration. Also secondary characteristics, such as properties of different acceleration components can be monitored.
There may be implemented several instructions for carrying out the detection of the key points. Examples of such instructions are given in the following list:
The above-listed instructions (and other such instructions) can be logically combined for enhancing the detection. For example, the start of the swing has been detected with good probability if two or three of the first listed requirements are met. On the other hand, the ending of the swing can only have happened if the swing has started and turned.
By means of the detection process disclosed above, it is possible to implement a swing monitoring program that does not need any input from the user during the monitoring session. However, there can be also implemented interactive monitoring programs or semi-interactive monitoring programs. In an exemplary semi-interactive program the device informs the golfer by a sound signal when it is ready for a new swing after it has detected that the player is in starting stance (i.e., when the device has been essentially motionless for a while).
The swing detection system can also utilize higher level artificial intelligence, such as fuzzy logic or learning systems, which adapt to a certain style of swinging and thus provide more reliable results. Detection can also be implemented by storing swing data temporarily and comparing the data with a pre-recorded reference swing acceleration profile or profiles in order to find similarities between them and to detect the key points that way.
In one embodiment, the signal given by the sensor can also be stored for further analysis by the device or by external data processing means, such as a computer. By this embodiment, the swings of the golfer can be analyzed thoroughly and/or developing of the swing of a golfer can be monitored in the long run in detail.
A timer unit 45 is used for obtaining correct time stamps for the key points of a swing. A timer unit can comprise a timer used commonly for performing typical timing functions of wrist watches, for example.
The memory unit 42 may be comprised of built-in memory, portable memory, or both.
The microprocessor 41 can be programmed to handle the signal processing needed in determining the key points of the swing and the calculations needed in determining the characteristic numbers. Alternatively, all or some of the processing and calculations can be performed using specialized electronic components, such as microchips. The raw of refined (extracted) data on the individual swings can also be transferred to separate data processing means, such as a computer, for calculation of the characteristic number(s).
In a preferred embodiment, all the steps needed for determining and reporting the characteristic number(s) are carried out in a single device. The device can be manufactured light and implemented as a wrist watch or wristop computer.
A test set of 10 swings of 12 different golfers was performed in order to illustrate the capabilities of the present device and method. Two swing properties were chosen to be monitored, namely the duration of the backswing (property 1) and the duration of the down-swing (property 2). Table 1 shows the mean values (μ) and standard deviations (σ) of the results. Also the handicaps of the golfers are shown.
A classification matrix based on the data on Table 1 is shown in Table 2. A classification function was used to classify the test persons into two groups based on their swing index numbers calculated from σ2 and σ2. The groups consisted of those having a hcp between 0 and 10 and of those having a hcp more than 10. The results show, that only one test person was classified into a wrong group with the classification function used. Thus, the proportion of correct observations in this case is about 92%, the error rate being about 8%.
The experimental results disclosed is this example illustrate the potential of the invention. It should be noticed that despite the low number of swings, namely 10, and only two parameters of interest used in the experiment, the derived swing index values correlate well with the hcp values of the players.
The average predicted hcp of the players was 28 (st. dev. 20), the average actual hcp being 25 (st. dev. 16). The correlation factor between the predicted and actual hcp values was 0.8. The correlation can be considered really high, taking into account that the number of repetition was only six.
This application is a Continuation of application Ser. No. 11/176,359 filed on Jul. 8, 2005 now U.S. Pat. No. 8,226,494. The entire contents of all of the above application is hereby incorporated by reference.
| Number | Name | Date | Kind |
|---|---|---|---|
| 3717857 | Evans | Feb 1973 | A |
| 3945606 | McDonald | Mar 1976 | A |
| 3945646 | Hammond | Mar 1976 | A |
| 4126314 | Martin | Nov 1978 | A |
| 4699379 | Chateau et al. | Oct 1987 | A |
| 4821218 | Potsch et al. | Apr 1989 | A |
| 4991850 | Wilhlem | Feb 1991 | A |
| 5154427 | Harlan et al. | Oct 1992 | A |
| 5233544 | Kobayashi | Aug 1993 | A |
| 5349853 | Oehler | Sep 1994 | A |
| 5375610 | LaCourse et al. | Dec 1994 | A |
| 5607361 | Mastandrea et al. | Mar 1997 | A |
| 5688183 | Sabatino et al. | Nov 1997 | A |
| 5694340 | Kim | Dec 1997 | A |
| 5792000 | Weber et al. | Aug 1998 | A |
| 5871406 | Worrell | Feb 1999 | A |
| 5907819 | Johnson | May 1999 | A |
| 6014887 | Naruo et al. | Jan 2000 | A |
| 6032530 | Hock | Mar 2000 | A |
| 6073086 | Marinelli | Jun 2000 | A |
| 6227984 | Blankenship | May 2001 | B1 |
| 6261102 | Dugan et al. | Jul 2001 | B1 |
| 6441745 | Gates | Aug 2002 | B1 |
| 6487906 | Hock | Dec 2002 | B1 |
| 6533675 | Funk | Mar 2003 | B2 |
| 6607450 | Hackman | Aug 2003 | B1 |
| 6648769 | Lee et al. | Nov 2003 | B2 |
| 6778866 | Bettwy | Aug 2004 | B1 |
| 6821211 | Otten et al. | Nov 2004 | B2 |
| 6856934 | Vock et al. | Feb 2005 | B2 |
| 6959259 | Vock et al. | Oct 2005 | B2 |
| 6991552 | Burke | Jan 2006 | B2 |
| 7022026 | Blankenship | Apr 2006 | B2 |
| 7033281 | Carnahan et al. | Apr 2006 | B2 |
| 7041014 | Wright et al. | May 2006 | B2 |
| 7101287 | Wagner | Sep 2006 | B1 |
| 7143639 | Gobush | Dec 2006 | B2 |
| 7166035 | Voges et al. | Jan 2007 | B2 |
| 7264554 | Bentley | Sep 2007 | B2 |
| 7271774 | Puurl | Sep 2007 | B2 |
| 7489241 | Mlettlnen et al. | Feb 2009 | B2 |
| 8123624 | Caldwell | Feb 2012 | B2 |
| 8616989 | Bentley | Dec 2013 | B2 |
| 20020187846 | Funk | Dec 2002 | A1 |
| 20030054327 | Evensen | Mar 2003 | A1 |
| 20030073518 | Marty et al. | Apr 2003 | A1 |
| 20030109322 | Funk et al. | Jun 2003 | A1 |
| 20040122538 | Gowan | Jun 2004 | A1 |
| 20040172213 | Kainulainen | Sep 2004 | A1 |
| 20050054457 | Eyestone et al. | Mar 2005 | A1 |
| 20050215335 | Marquardt | Sep 2005 | A1 |
| 20050261073 | Farrington et al. | Nov 2005 | A1 |
| 20050282650 | Miettinen et al. | Dec 2005 | A1 |
| 20060199659 | Caldwell | Sep 2006 | A1 |
| 20070010341 | Miettinen et al. | Jan 2007 | A1 |
| 20070091004 | Puurl | Apr 2007 | A1 |
| 20140114453 | Bentley | Apr 2014 | A1 |
| Number | Date | Country |
|---|---|---|
| 05-023409 | Feb 1993 | JP |
| WO 0235184 | May 2002 | WO |
| WO 2005094949 | Oct 2005 | WO |
| Entry |
|---|
| “Standard Normal Scores / Z Scores / Z values”, sourced from http://www.stats.uwo.ca/ugstudies/FallWinter99/hs201nov10.htm, (Nov. 10, 1999). |
| Number | Date | Country | |
|---|---|---|---|
| 20120264534 A1 | Oct 2012 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 11176359 | Jul 2005 | US |
| Child | 13532476 | US |
