Racket sports are those that involve the use of a racket or paddle to hit a ball or another object over a net or into a designated area of play.1 The main objective in these sports is to hit the object in such a way that the opponent is unable to return it. Today, badminton, tennis, and table tennis stand out as some of the most widely practiced sports globally.2 These sports demand the execution of a substantial number of strokes and involve short movements with frequent changes of direction.3,4 Given their popularity2 and the increasing physical demands they impose,3,4 sports scientists have developed a growing interest in conducting applied research on athletic training in racket sports. One area that has gathered significant attention in recent times is that of training load and its management.5–7 Although the term’s meaning may be ambiguous,8 it generally refers to the combination of exercise volume and intensity.9 Training and competition loads can be classified as either external or internal.10,11 External load refers to the amount of work that an athlete performs, and internal load corresponds to the athlete’s psychophysiological response to the external load.12 In this respect, it is considered that the quantification of internal and external load parameters is essential to the correct load management and design of training programs in racket sports.7,9–11,13 Moreover, effective control and management of training and competition internal and external loads have been shown to be critical in achieving desired performance improvements and to minimize the risk of injury in multiple sports involving striking implements.6,7,14,15
The repetitive nature of stroke performance in racket sports highlights the importance of their external load quantification. As external load metrics require specificity,10 hitting load is recognized as a primary metric to be measured in this regard.16,17 This feature sets racket sports apart from other team or situational sports as a result of its complex motor skills that require coordination across the player’s kinetic chain to achieve optimal performance.18,19 Specifically, stroke performance efficiency depends on various factors, such as speed, angle of impact, spin direction, and precision.20,21 Forces must also be coordinated from the ground up through the kinetic chain to upper limbs and, finally, to the racket, utilizing elastic energy and muscle prestretch.20,22,23 For instance, recently, it has been shown that the capability to generate force in short periods (<250 ms) is relevant to develop high serve velocities in competitive tennis players.24 Moreover, in combination with other factors, repetition of high biomechanical stresses caused by tennis strokes may play an important role in upper limb injuries16,25,26 and low back pain.27 Research in other overhead sports has delved into the relationship between workload and injury. Dennis et al15 identified an optimal number of throws in cricket practice per week as injury risk rises with both higher and lower repetitions than this optimal threshold. So, quantifying an optimal range for hitting volume in racket sports could help coaches to optimize training programs and to minimize overuse injuries.16,17,28
In recent years, advancements in technology such as global technology systems, inertial measurement units (IMUs), and camera systems have enabled new ways to monitor external load in multiple sports.13,29–31 These technologies are particularly useful for measuring gross movement metrics, such as distances, accelerations and decelerations, impacts, and metabolic power, which are key determinants of external load in many team sports.32,33 In racket sports, movement load demands have also been analyzed during competition using optical tracking systems, such as Hawk-Eye34,35 and Foxtenn.21 Despite the literature advancements in quantifying movement loads during competition,34,35 there have been only a few studies that have specifically reported on hitting loads in racket sports.34–37 In tennis, the hitting volume of Australian Open players was monitored, showing that the average number of ground strokes per match was 312 (max = 1292) for males and 192 (max = 391) for females.34 Likewise, Whiteside and Reid35 reported male and female players’ cumulative strokes played across the first 4 rounds of Australian Open tournaments from 2012 to 2016, with a mean of 1,49,629 (SD: 20,546) and 2,28,139 (SD: 50,707), respectively. In the case of other racket sports, hitting load has been reported in elite badminton,36 squash,37 and padel38 players.
Although some studies have quantified the hitting load in high-level athletes, stroke quantification is typically not monitored at the either formative or competitive level.5 In addition, current hitting load quantification methods in racket sports are expensive (i.e., Hawk-Eye and Foxtenn), impractical (i.e., observational tools), or inaccurate (subjective ratings), leading to the intuitive control of the training load by most tennis coaches.39 However, monitoring the hitting load is essential for the accurate quantification of training and competition loads in racket sports players of all levels and ages.16 Despite the existence of scientific reviews on new technologies in racket sports,40 none has been conducted yet on the commercialized racket and arm-mounted devices that can monitor the hitting load in racket sports. Therefore, the aim of this study was to perform a scoping review following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) extension for Scoping Reviews guidelines to identify the primary racket and arm-mounted technologies that enable the quantification of the hitting load in racket sports. In addition, the study assessed the reliability of measurements obtained from these sensors.
Methods
Experimental Approach to the Problem
This work was conducted following the PRISMA extension for Scoping Reviews statement.41,42
Eligibility Criteria
To be eligible for this review, the records had to (1) be published in English, (2) be conducted on human participants, and (3) use racket or arm-mounted inertial sensors to quantify the hitting load in at least 1 racket sport. The English language aimed to enhance accessibility to a broader audience. Given our primary objective of identifying current technologies in the sports market, we did not impose restrictions regarding the players’ skill level, study designs, outcomes of interest, or interventions.
Records were excluded if they did not primarily focus on monitoring hitting load or if they utilized inertial sensors that had not yet been commercialized. The exclusion of uncommercialized sensors, such as prototypes, aimed to ensure the practical applicability of our research. In addition, reviews, theses, and conference papers were also excluded, apart from conference papers on IEEE Xplore search.
Sources of Information and Strategy
The same Boolean terms and keywords were used for an electronic literature search across various databases, such as PubMed, SPORTDiscus, Web of Science, and IEEE Xplore, in all fields (“Racket sports” OR “Tennis” OR “Badminton” OR “Squash” OR “Padel” OR “Pickleball” OR “Beach Tennis” OR “Touch Tennis” OR “Frontenis” OR “Racquetball”) AND (“Wearable electronic device” OR “Wearable sensor” OR “Inertial Measurement Unit” OR “IMU” OR “Inertial sensor” OR “Accelerometer” OR “Racket sensor” OR “Blade sensor”). Simultaneously, a gray literature search was conducted via Google Search Engines using the same keywords in all fields and following the guidelines provided by Godin et al.43
The final searches were conducted on February 28, 2023.
Data Extraction
The authors collectively discussed and agreed upon the variables collected from the final selected studies. The first author was responsible for data extraction, and the chosen variables were categorized into 2 groups: sensor characteristics and parameters of analysis. First, sensor characteristics variables, such as system’s type and location, brand, model, cost, weight, size, battery life, connection protocol, and international approval or any other use limitations, were included. The system type indicated the specific sport for which the sensor was designed, and the system location referred to the device’s positioning during use.
Second, the data related to parameters of analysis, which described the parameters tracked by the sensors, were classified into the following categories: external load volume-related and load intensity-related variables, internal load variables (e.g., heart rate [HR] and heart rate variability [HRV]), and hitting quality variables (Table 1).
Description of External-Load and Hitting-Quality Variables
| External-load variables | Volume-related variables | Hitting volume | N of strokes | The device records the total number of strokes hit without distinguishing the type |
| N of each stroke | The device records the total number of strokes hit and distinguishes between all stroke types, including FH, BH, serve, FHV, BHV, and smash | |||
| N of FH | The device records the total number of FH hit | |||
| N of BH | The device records the total number of BH hit | |||
| N of ground strokes | The device tracks the total number of ground strokes hit, which includes FH and BH | |||
| N of serves | The device records the total number of serves hit | |||
| N of FHV and BHV | The device records the total number of FHV and BHV hit | |||
| N of smash | The device records the total number of smashes hit | |||
| N of offensive and defensive strokes | The device can distinguish between overhead and abovehead strokes and record the total number of offensive and defensive strokes | |||
| Duration | Total time | The device measures the complete activity duration | ||
| Effective playing time | The device measures the time spent playing | |||
| Intensity-related variables | Hitting velocity | Stroke’s speed, km/h | The velocity at which the ball is traveling through the air after any stroke | |
| Stroke’s acceleration, m/s2 | The rate at which player’s racket increases in speed from the start of the swing to the point of contact with the ball | |||
| Swing speed, km/h | The maximum speed at which a player’s racket is moving during the forward swing of a stroke | |||
| Serve speed, km/h | The velocity at which a player’s serve travels after it is hit from the racket | |||
| Hitting frequency | Rally quality | The average number of strokes per rally | ||
| Match pace | The average number of strokes played per minute during rallies | |||
| Hitting-quality variables | Ball spin | The rotation of the tennis ball as its travel through the air after being hit by the player’s racket | ||
| Sweet spot | The area of the racket that is most effective at hitting the ball with maximum power, accuracy, and control | |||
| Swing motion | The movement of the player’s arm and racket as they prepare to hit the ball and follow through after making contact | |||
| Contact penalty | The quantity of vibrations that the racket transfers to the arm during the strokes | |||
| Heaviness | Estimation of how “heavy” the stroke is by combining stroke’s speed and spin | |||
| Serve type | The device can distinguish between topspin, slice, and flat serves | |||
| Spin identification | The device can distinguish between topspin, backspin, and flat strokes | |||
| Radian, ° | Describes the angle of strokes | |||
| PIQ score | System to measure tennis performance metrics, such as power, spin, and stroke type | |||
Abbreviations: BH, backhand; BHV, BH volley; FH, forehand; FHV, FH volley; N, number; PIQ, Performance Index Quotient.
The variables related to external load volume encompassed hitting volume and duration. Hitting volume specifications included whether the total number of strokes performed could be recorded or whether they could be distinguished by types. In instances where reliability studies were available, details such as sample size, protocol, and results in terms of intraclass correlation coefficient (ICC), percentage of agreement, and Cohen kappa coefficient (κ) were provided. Regarding duration, it was indicated whether practice time and/or effective playing time could be recorded by the device. The ICC values were interpreted as follows: excellent (>.85), high (.75–.85), moderate (.40–.75), and poor (<.40).
The variables related to the external load intensity included hitting velocity and hitting frequency. Specifications for hitting velocity entailed details about the types of strokes it could record, accompanied by reliability studies, such as percentage of deviation, Pearson correlation coefficient (r), or ICC, if applicable.
For internal load, specifications were provided regarding whether it could be recorded in various ways by the devices, such as HR and HRV. Concerning hitting quality variables, it was specified whether the impact on the sweet point of the racket and the ball spin could be identified by the sensors, accompanied by reliability analysis, if available.
Results
Study Selection
Figure 1 uses the PRISMA extension for Scoping Reviews flowchart to summarize all the stages of the selection process. Although the first author was responsible for the selection and screening process, any discrepancies were solved through consensus after a thorough review of the wearables that caused controversy. During the initial stage, 484 potentially relevant articles were identified from various databases, including PubMed (n = 101), SPORTDiscus (n = 56), Web of Science (n = 231), and IEEE Xplore (n = 69). Upon further screening, a total of 162 duplicated records were identified, and 274 records were eliminated after reading the title. Therefore, 48 records that met our initial screening criteria were selected. During the eligibility assessment, 29 records that did not meet the eligibility criteria were excluded. Finally, after thorough review, a total of 19 records were considered suitable for the scoping review. Specifically, 7 records were obtained from scientific databases and 12 from gray literature.
—PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) extension for Scoping Reviews flowchart.
Citation: International Journal of Sports Physiology and Performance 19, 6; 10.1123/ijspp.2023-0385
Characteristics of Sensors
Nineteen different devices were identified and included in the study. Among these, 11 were specifically designed for tennis (Table 2), 4 for badminton, 2 for table tennis, 1 for squash, and 1 for padel (Table 3). None of these devices was compatible across multiple racket sports.
Characteristics of Tennis Sensors
| Location | Brand | Model | Commercial status and cost, $ | Weight and size | Battery life, h | Connection protocol | ITF approval and other limitations |
|---|---|---|---|---|---|---|---|
| Wrist | Armbeep (Slovenia) | Armbeep Tracker44 | Out of stock $599 (bracelet) or $149 (Apple watch) | 12 g 34 × 29 × 12 mm3 or Apple Watch | 6 | Bluetooth to smartphone or “Apple Watch” app | Approved by ITF42 No limitations |
| Babolat (France) | Babolat Pop45 | Discontinued (March 21, 2021) | 9 g NA | 6 | Bluetooth to smartphone app | Approved by ITF42 No limitations | |
| Artengo (France) | Artengo Personal Coach46 | Discontinued (2021) | 22.8 g NA | 6 | Sensor to Artengo watch by ANT + Protocol Artengo watch to computer by USB. | Approved by ITF42 No limitations | |
| Grip-embedded | Sony (Japan) | Sony Smart Sensor47 | Discontinued (September 30, 2021) | 7.9 g 31.3 × 31.3 × 17.6 mm3 | 3 | Bluetooth to smartphone app | Approved by ITF42 Compatible with most of Wilson, Head, Prince, Yonex, and Srixon rackets |
| Head (Austria—United States) | Head Zepp48 | Discontinued (2021) | 7 g NA | 5–6 | Bluetooth to smartphone app | Approved by ITF42 Compatible with most of “Head” rackets | |
| Babolat (France) | Babolat Play49 | Discontinued (March 21, 2021) | 20 g 10 × 10 × 40 mm3 | 6–7 | Bluetooth to smartphone app | Approved by ITF42 Integrated in few Babolat rackets | |
| Grip-attached | Zepp (China) | Zepp Tennis 1.050 | Discontinued (2017) | 7.7 g 28 × 28 × 11 mm | 4 | Bluetooth to smartphone app | Approved by ITF42 No limitations |
| Head (Austria—USA) | Zepp Tennis 2.051 | Out of stock $99 | 6.25 g 25.4 × 25.4 × 12.3 mm3 | 8 | Bluetooth to smartphone app | Approved by ITF42 No limitations | |
| Coollang (China) | Coollang Sensor52 | $129 | 6 g 30 × 30 × 10 mm3 | 6 | Bluetooth to smartphone app | Not approved by ITF42 −5 °C to 50 °C | |
| Dampener | Babolat (France) | Courtmatics53 | $99 | 6 g 20.3 × 20.3 × 13 mm3 | 5–6 | Bluetooth to smartphone app | Not approved by ITF42 Only in wet conditions |
| Qlipp (Singapore) | Qlipp54 | $70 | 8 g 30.3 × 26.7 × 10.7 mm3 | 4–6 | Bluetooth to smartphone app | Not approved by ITF42 Only in wet conditions |
Abbreviations: ITF, International Tennis Federation; NA, not available; ANT, adaptive network topology.
Characteristics of Other Racket-Sports Sensors
| Sport | Location | Brand | Model | Commercial status and cost | Weight and size | Battery life, h | Connection protocol | International federation approval and other limitations |
|---|---|---|---|---|---|---|---|---|
| Badminton | Grip attached | Collang (China) | Smart Xiaoyu 255 | Discontinued | 5 g 30 × 20 × 8.7 mm3 | 10 | Bluetooth to smartphone app | NA |
| Coollang (China) | Koospur Xiaoyu 356 | $59 | 5 g 27.2 × 23.9 × 9.3 mm3 | 10 | Bluetooth to smartphone app | NA | ||
| Actofit (India) | Actofit Badminton Tracker57 | Out of stock $73 | 6 g 29.9 × 22 × 9.9 mm3 | 10 | Bluetooth to smartphone app | NA | ||
| Usense (China) | Usense 1.058 | Discontinued $49.90 | 8 g 28 × 28 × 12 mm | 3 | Bluetooth to smartphone app | NA | ||
| Squash | Grip attached | Racketware (England) | Racketware 1.059 | $182.80 | 9.5 g 24 × 24 × 13 mm3 | 10 | Bluetooth to smartphone app | PSA partner NA |
| Padel | Grip embedded | Kaitt (Spain) | Kaitt Unik60 | Discontinued $369 | 365 g (blade) NR | NA | Bluetooth to smartphone app or blade display | NA Only with Kaitt Unik Blade |
| Table tennis | Grip embedded | GogoTak (South Korea) | Choreiking 161 | NA | 86.2 g (blade) NR | 3 | Bluetooth to smartphone app | Not approved by ITTF Only with Choreiking 1 Blade |
| GogoTak (South Korea) | Choreiking 243 | NA | 86.2 g (blade) NR | 5 | Bluetooth to smartphone app | Approved by ITTF Only with Choreiking 2 Blade |
Abbreviations: ITTF, International Tennis Table Federation; NA, not available; NR, not reported; PSA, Professional Squash Association.
Four different locations were identified: wrist systems, which are wearable devices intended to be worn on the wrist; grip-attached systems, which are devices affixed to the racket’s grip and alter its swing weight; grip-embedded systems, which are devices integrated or embedded into the racket without modifying its swing weight; and dampener systems, which are devices attached to the racket’s strings. The most prevalent type of system was grip-attached sensors (n = 8), accounting for 42.1% of the devices. This was followed by grip-embedded sensors (n = 6; 31.6%), wrist sensors (n = 3; 15.8%), and dampener sensors (n = 2; 10.5%).
The development of the 19 sensors identified has been undertaken in various countries, with significant contributions from China (n = 5; 26.3%); France (n = 4; 21.1%); South Korea (n = 2; 10.5%); Austria and the United States (n = 2; 10.5%); and Slovenia, Japan, Singapore, and India (n = 1; 5.3%). The devices ranged in weights between 5 and 22.8 g, excluding the grip-embedded devices that accounted for the total weight of the implement. The battery life of the sensors ranged between 2 and 10 hours. All the devices transferred data via Bluetooth except for the Artengo Personal Coach, which required a USB (universal serial bus) cable. Regarding their current commercial status, it was found to be quite complex, with most of them either discontinued or facing distribution issues (n = 12; 73.7%). Their commercialization started in 2013 and experienced fluctuations until 2022, after which no further systems have been developed.
Among the tennis devices (n = 11), there were 3 wrist sensors, 3 grip-attached sensors, 3 grip-embedded sensors, and 2 dampener sensors (Table 2). Specifically, the availability of tennis sensors had shown a positive trend until 2020, with a total of 10 devices being available. In 2021, half of these devices were discontinued, although the remaining 5 devices have continued to be available to the present (Figure 2).
—Number of tennis sensors commercially available from 2010 to 2023.
Citation: International Journal of Sports Physiology and Performance 19, 6; 10.1123/ijspp.2023-0385
Most of tennis sensors analyzed in this study (n = 8; 72.7%) have been approved by the International Tennis Federation and are suitable for use in official competitions.44 Conversely, most of sensors designed for other racket sports (n = 7; 87.5%) have not been approved by their respective federations, or the information is not available. One exception is the Choreiking 245 sensor, which has been approved by the International Table Tennis Federation.
Parameters of Analysis of Sensors
Tables 4 and 5 provide a summary of the parameters of analysis monitored by the sensors reviewed, which are compatible with tennis and other racket sports, respectively. The reliability studies included samples of amateur players (n = 1; 14.3%) or competitive players (n = 6; 85.7%). The protocol for these studies involved analytic drills (n = 1; 14.3%), simulated match play (n = 3; 42.9%), or a combination of both methodologies (n = 3; 42.9%).
Parameters of Analysis of Tennis Sensors
| Model | External load | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Volume | Intensity | |||||||||||
| Hitting volume | Sample and protocol | R | Duration | Hitting velocity | Reliability | HF | Reliability | Internal load | Reliability | Hitting quality | Reliability | |
| Armbeep Tracker44 | N of each stroke | Competitive players on simulated MP | ICC of N of strokes62: .976 95% CI, .945–.989 | Total time | Stroke speed and acceleration | NR | N of strokes per period | NR | HR and HRV | NR | Contact penalty | NR |
| Babolat Pop45 | N of each stroke | Competitive players on simulated MP | % Agr of N of FH, BH, and S63: 78.8%, 89.9%, and 69.9% | Total time and effective playing time | Serve speed63 | % Dev: 2.9%–15.9% | NR | NR | NR | NR | Ball spin and PIQ score | NR |
| Stroke speed and swing speed | NR | |||||||||||
| Artengo Personal Coach46 | N of each stroke | NR | NR | Total time and effective playing time | Serve speed | NR | N of strokes per period | NR | HR (with heart belt) and EE | NR | Sweet spot | NR |
| Sony Smart Sensor47 | N of each stroke | Competitive players during 6-wk training block (including analytic drills and MP) | % Agr of N of strokes64: 94% | Total time and effective playing time | Stroke speed and swing speed | NR | NR | NR | NR | NR | Sweet spot and ball spin | NR |
| % Agr of FH, BH, S, FHV, BHV, and Smash64: 99%, 96%, 95%, 86%, 82%, and 70% | ||||||||||||
| % Agr grouping into FHS, BHS, and OHS64: 96%, 91%, and 92% | ||||||||||||
| Head Zepp48 | N of each stroke | Competitive players on simulated MP | % Agr of N of FH, BH, and S63: 80.8%, 90.3%, and 97.0% | Effective playing time | Serve speed63 | % Dev: 3.4%–6.7% | NR | NR | NR | NR | Sweet spot and ball spin | NR |
| Stroke speed | NR | |||||||||||
| Babolat Play49 | N of each stroke | Competitive player on analytic drill | % Agr of N of strokes12: 100% | Total time and effective playing time | Serve speed63 | % Dev: 0%–19.6% | NR | NR | NR | NR | Sweet spot12 | κ = .412; 95% CI, .075–.749 |
| Competitive players on simulated MP | % Agr of N of strokes12: 99.33% | |||||||||||
| Competitive player on analytic drill | κ of N of each stroke12: 0.730 95% CI, .541–.920 | |||||||||||
| Competitive players on analytic drill | % Agr of N of groundstrokes63: 98.75% | Stroke speed12 | r = .708; P < .001 | Ball spin | NR | |||||||
| Competitive players on simulated MP | % Agr of N of groundstrokes63: 98.26% | |||||||||||
| Competitive players on simulated MP | % Agr of N of FH, BH, and S63: 83.0%, 93.9%, and 95.6% | |||||||||||
| Zepp Tennis 1.050 | N of each stroke | NR | NR | Total time | Stroke speed | NR | NR | NR | NR | NR | Sweet spot and ball spin | NR |
| Zepp Tennis 2.051 | N of each stroke | Competitive player on analytic drill | % Agr of N of strokes12: 100% | Total time and effective playing time | Swing speed12 | ICC: .983; P < .001 | NR | NR | EE, kcal | NR | Sweet spot12 | κ = .257; 95% CI, .099–.613 |
| Competitive player on simulated MP | % Agr of N of strokes12: 96.67% | Stroke speed65 | r = .83 | |||||||||
| Competitive player on analytic drill | κ of N of each stroke12: 0.612 95% CI, .391–.834 | Serve speed | NR | Ball spin | NR | |||||||
| Coollang Sensor52 | N of each stroke | NR | NR | Total time and effective playing time | Stroke speed and swing speed | NR | N of strokes per minute | NR | EE, kcal | NR | Ball spin | NR |
| Courtmatics53 | N of each stroke | NR | NR | NR | Serve speed | NR | NR | NR | EE, kcal | NR | Sweet spot, ball spin, serve type and swing motion | NR |
| Qlipp54 | N of each stroke | NR | NR | NR | Stroke speed65 | r = .66 | NR | NR | NR | NR | Sweet spot, spin identification, and heaviness | NR |
| Stroke acceleration | NR | |||||||||||
Abbreviations: % Agr, percentage of agreement; BH, backhand; BHS, BH swing; BV, BH volley; % Dev, percentage of deviation; EE, energy expenditure; FH, forehand; FHS, FH swing; FV, FH volley; HF, hitting frequency; HR, heart rate; HRV, heart-rate variability; ICC, intraclass correlation coefficient; κ, Cohen kappa coefficient; Lab, laboratory setting; MP, match play; N, number; NR, not reported; OHS, overhead swing; r, Pearson correlation coefficient; S, serve.
Parameters of Analysis of Other Racket-Sports Sensors
| Type | System | External load | Internal load | Reliability | Hitting quality | Reliability | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Volume | Intensity | ||||||||||||
| Hitting volume | Sample and protocol | Reliability | Duration | Hitting velocity | Reliability | HF | Reliability | ||||||
| Badminton | Smart Xiaoyu 21 | N3 of each stroke and N of offensive and defensive strokes | NR | NR | Total time | Stroke power and speed | NR | NR | NR | EE, kcal | NR | Radian, ° | NR |
| Koospur Xiaoyu 32 | N of each stroke and N of offensive and defensive strokes | NR | NR | Total time | Stroke power and speed | NR | NR | NR | EE, kcal | NR | Radian, ° | NR | |
| Actofit Badminton Tracker3 | N of each stroke and N of offensive and defensive strokes | NR | NR | Total time | Stroke power and speed | NR | NR | NR | EE, kcal | NR | Radian, ° | NR | |
| Usense 1.04 | N of each stroke and N of offensive and defensive strokes | Amateur players on analytic drill | % Agr on stroke counting5: 79.32% | Total time | Stroke speed | NR | NR | NR | EE, kcal | NR | Sweet spot | NR | |
| Squash | Racketware 1.06 | N of strokes | NR | NR | Total time | Stroke power | NR | Match pace and rally quality | NR | EE, kcal and Effort Level | NR | Swing motion | NR |
| Padel | Kaitt Unik7 | N of each stroke | NR | NR | Total time and effective playing time | Stroke speed | NR | NR | NR | EE, kcal | NR | Radian, ° | NR |
| Table tennis | Choreiking 18 | N of each stroke | NR | NR | Total time | Swing speed | NR | NR | NR | NR | NR | Radian, °, and sweet spot | NR |
| Choreiking 29 | N of each stroke | NR | NR | Total time | Swing speed | NR | NR | NR | NR | NR | Radian, °, sweet spot and swing motion | NR | |
Abbreviations: % Agr, percentage of agreement; EE, energy expenditure; HF, hitting frequency; N, number; NR, not reported.
External-Load Volume-Related Variables
All the sensors reviewed in this study were capable of quantifying hitting volume in at least 1 racket sport. So, most of these sensors could measure the number of each type of stroke performed during the practice sessions. However, 1 exception was the Racketware sensor,46 which could only provide the total number of strokes without differentiation between types. It is important to highlight that, although included in their technical specifications, only 7 of these sensors conducted a validity and reliability analysis specifically for hitting volume tracking.
Regarding the tracking of the total number of shots, the Sony Smart Sensor47 exhibited a high reliability rate of 94%,48 and the Armbeep Tennis Tracker49 showed an ICC of .976.50 The Babolat Play51 and Zepp Tennis 2.052 achieved 100% reliability in lab settings12 but were less reliable in matches (99.33% and 96.67%, respectively). The Babolat Play51 and the Zepp Tennis 2.052 achieved a reliability rate of 99.3% and 96.7%, respectively.12
However, sensors were found to be imprecise in distinguishing between different types of strokes, as reported in several studies.12,48,50,53,54 Notably, the Sony Smart Sensor47 exhibited strong correlations when categorizing strokes into forehand (forehand and forehand volleys), backhand (backhand and backhand volleys), and overhead strokes (serve and smash), with percentages of 96%, 91%, and 92%, respectively.48 Similarly, the Babolat Play51 sensor showed excellent reliability in tracking ground strokes in both lab and match play settings, with success rates of 98.8% and 98.3%, respectively.53
In addition to monitoring hitting volume, most of the sensors (n = 17; 89.5%) could track the duration of the activity performed. Furthermore, although 7 of those sensors were able to track the effective playing time during sessions, none has been scientifically validated. Notably, the tennis dampener sensors55,56 did not provide any information regarding the duration of practice.
External-Load Intensity-Related Variables
First, it is worth noting that all the sensors included in this review provided information about the hitting velocity of the strokes played. However, Artengo Personal Coach57 and Courtmatics55 sensors were only capable of providing such information about the serve. Second, several sensors on the market (21.05%), such as the Armbeep Tracker,49 Artengo Personal Coach,57 Coollang Sensor,58 and Racketware,46 recorded data related to hitting frequency, such as the number of strokes per time period, rally quality, and match pace. However, despite their technical specifications, none of the sensors included in this review had reported successful scientific validation to measure external load intensity-related variables.
Internal-Load Variables
In addition, despite their technical specifications, none of the 19 sensors has been scientifically validated to measure internal load variables. Out of the sensors reviewed, 10 were able to display calorie consumption but with questionable reliability. Furthermore, only the Armbeep Tracker49 quantified HR and HRV during practice. In addition, the Artengo Personal Coach57 app had the functionality to analyze HR, but an additional HR belt was required.
Hitting-Quality Variables
Among the tennis sensors reviewed, the most frequently quantified variables were ball spin (n = 9; 81.8%) and the point of impact on the racket head (n = 8; 72.7%). In contrast, spin was not measured by any of the other racket sports sensors, and only half of the devices (4 out of 8) provided data on the point of impact location during strokes. Among the variables frequently measured by these devices, the angle of impact was the most common, with 6 sensors capable of measuring it. Interestingly, none of the hitting quality measures generated by the sensors have been scientifically validated.12
Discussion
The aim of this scoping review was to identify the primary racket and arm-mounted IMU technologies that enable hitting load quantification in racket sports. Despite the existence of multiple devices in the sports market, many of them exhibit relevant limitations, lack scientific validation, or face commercial challenges. As a result, despite the considerable efforts that have been made by the manufacturers, the existing racket and arm-mounted solutions within the sports industry do not adequately address the need to effectively quantify hitting load in racket sports.
Accurate management of training and competition loads is widely recognized as crucial for enhancing performance outcomes and minimizing the risk of injury across sports involving striking implements.6,7,15 Scientific literature emphasizes the importance of accurately quantifying both internal and external load parameters to effectively manage training loads and to design appropriate training programs.7,9–11,13 In the context of racket sports, hitting volume, in addition to movement loads, is considered a fundamental external load variable to consider.16,34,35 Therefore, a sensor capable of quantifying hitting volume, preferably along with other external and internal load parameters, could be a valuable tool for tennis coaches and specialists. In this context, the advancement of prototype machine learning algorithms has enabled reliable quantification of hitting volume, shot classification, and other internal and external load variables using a cervically mounted wearable sensor (Catapult OptimEye S5, Catapult).66 However, these systems are characterized by their complexity and a significant price point, which can present challenges for tennis specialists. In recent years, the IMU sports market has seen notable advancements67 with the introduction of more user-friendly and cost-effective alternatives. This includes the commercialization of at least 19 racket and arm-mounted sensors designed to quantify hitting load in racket sports. Among these sensors, the majority are specifically designed for tennis (n = 11), although a limited number have been developed for badminton, table tennis, squash, and padel (n = 4, 2, 1, and 1, respectively). Within the realm of tennis, various sensor types have been commercialized, including 3 grip-attached, grip-embedded, and wrist sensors along with 2 dampener sensors. In contrast, other racket sports feature a single kind of sensor, such as grip-attached sensors for badminton and squash or grip-embedded sensors for table tennis and padel. Unfortunately, no sensors were found for several racket sports, including pickleball, beach tennis, touch tennis, frontenis, or racquetball. The greater sensor penetration in tennis broadly mirrors the larger amount of applied research activity in tennis compared with other racket sports. Despite the lack of validation, many of these sensors offer additional functionality for monitoring other external and internal load variables as well as hitting quality.
Among the tennis sensors, only 6 out of the 11 have undergone a validation process for hitting volume quantification.12,48,50,53 Nevertheless, only the Babolat Play,51 Sony Smart Sensor,47 Zepp 2.0,52 and Armbeep Tracker49 have shown excellent reliability (>.85) in monitoring hitting volume, in laboratory settings, and/or in real match play conditions.12,48,50,53 It is important to note that our review did not identify significant differences in reliability based on the types of devices. Although these 4 sensors can accurately measure the total number of strokes performed, they struggle to differentiate between different stroke types.12,48,50,53 However, there are 2 devices that provide more accurate information beyond just the total number of strokes. First, the Sony Smart Sensor47 has shown reliability by categorizing the strokes into forehand, backhand, and overhead strokes.48 Similarly, the Babolat Play sensor51 has demonstrated reliability by specifically analyzing the number of ground strokes.53 Nonetheless, both options have been discontinued. Thus, Armbeep Tracker49 and Zepp 2.052 are the only wrist and racket-mounted options currently available in the sports market for a reliable quantification of the hitting volume, despite the challenges associated with their commercial distribution. On the other hand, none of the other racket sports sensors have undergone reliable validation for hitting volume quantification.68 Notably, only the Usense 1.0 sensor65 has been subjected to scientific validation but did not yield successful results.54 Therefore, it seems that the sports industry currently lacks the necessary tools to effectively address hitting load monitoring needs in racket sports.
In addition to hitting volume, some sensors purport to monitor external load parameters related to the practice’s intensity, such as hitting velocity and hitting frequency. These parameters are not only valuable for load management but also have a significant impact on enhancing performance in racket sports.34,69 Studies have assessed the reliability of these sensors for hitting velocity,12,53,64 and Zepp 2.052 has shown promising results, exhibiting strong agreement with swing speed.12 However, there is a lack of empirical work validating measures of hitting frequency. Therefore, Zepp 2.052 is currently the only commercially available racket-mounted sensor that reliably enables the measurement of both volume and intensity-related external load parameters. Furthermore, several sensors provide internal load variables, with energy expenditure being the most identified variable. However, no validation studies have been conducted on these sensors, and the reliability of measuring calorie consumption solely without considering other internal load variables remains questionable. In contrast, the Armbeep Tracker49 is the only arm-mounted sensor capable of independently capturing internal load values, such as HR and HRV. Thus, as well as the cervically mounted sensor mentioned earlier,66 the Armbeep Tracker49 is unique in reliably monitoring both external hitting load and internal load variables for tennis players. Finally, most sensors provide parameters related to hitting quality, such as sweet spot and ball spin, according to their technical specifications. However, these parameters have not been reliably validated in the literature.12
In addition to the variables tracked by the devices, it is important to emphasize that an ideal sensor should be nonintrusive and should not impact the functionality of the implement. Surprisingly, among the sensors examined, racket-embedded sensors were the only ones that fully met these criteria, providing a convenient and unobtrusive solution. Several limitations were identified with other type of sensors. First, racket-attached sensors have been found to alter the swing weight of the racket, affecting the player’s swing technique.70 Similarly, dampener sensors, attached to the racket strings, could introduce similar limitations due to their weight (6–8 g). These characteristics inherently limit the usability of these sensors, often depending on the athlete’s skill level. In addition, wearing a sensor attached to the wrist, which plays a significant role in strokes,71 could cause discomfort for players. However, recent literature has not found any negative effects on velocity production or accuracy when wearing heavier wrist weights (50–200 g).72 Although racket-embedded sensors offer greater comfort during play compared with other types of sensors, they also have limitations. One of these limitations is their compatibility as they are specifically designed for certain racket models. For example, Babolat Play,51 Kaitt Unik,73 Choreiking 1,74 and Choireiking 245 sensors are integrated into their corresponding racket models. The Zepp Head60 sensor is only compatible with selected Head rackets. Sony Smart Sensor47 appears to be less restrictive as it can be embedded with a wider range of racket models.
Considering the aforementioned factors, it seems evident that the sports industry faces limitations not only in terms of the reliability of the variability monitored by the devices but also in enhancing the comfort and compatibility of these devices. By addressing these aspects, the industry could ensure the effective use of these devices by both elite players and those in formative stages. The limited availability of reliable, comfortable, and economically accessible options for load monitoring extends beyond the racket sports market. The support provided by the sports industry for other sports involving striking implements is even more limited. Although there are devices available for identifying swing motion and hitting quality parameters in sports like baseball,75,76 softball,75,76 golf,77–79 and ice hockey,80 there is currently no specific device designed for monitoring hitting load. However, research studies have explored prototypes that successfully quantify hitting volume in golf81 and hockey.82,83 In contrast, the situation differs significantly in team sports as there is a wide array of validated devices readily accessible for monitoring both external and internal load parameters.84–88 Moreover, although the sensors found for racket sports could only monitor sport-specific characteristics, most of the devices designed for team sports can measure external load parameters in multiple sports. Therefore, the sports industry should prioritize the development of unobtrusive and affordable solutions capable of quantify hitting load in racket sports while also ensuring compatibility across various racket sports, similar to the approach observed in team sports.
Practical Applications
Strength and conditioning coaches should prioritize the quantification of hitting load to enhance the performance of racket sports athletes across all levels. Consequently, technological advancements are essential to improve workload monitoring in these sports. In this context, optimal hitting volume could be established based on the athlete’s level, age, and/or gender. The development of sensors capable of monitoring hitting load holds broader implications beyond optimizing athletic training in racket sports. For instance, it could greatly assist in controlling and gradually increasing hitting loads during the return-to-play phases of injury rehabilitation for racket sports. Moreover, from an academic perspective, it would facilitate data collection for future applied research on the effects of hitting load on racket sports.
Conclusions
Accurate quantification of hitting load seems essential for optimal workload management in racket sports for both competitive and formative players.16,17 To effectively monitor hitting load, the sports industry should strive to provide convenient and accessible systems for quantifying hitting volume. Although some progress has been made in this area in recent years, there is still a lack of scientific validation, and the availability of such systems in the sports market remains limited. In tennis, only a few devices that enable reliable quantification of hitting load are currently available, along with other variables of external and internal load. In addition, in other racket sports, the availability of reliable devices for quantifying hitting load is significantly more limited.
References
- 1.↑
O’Donoghue P, Reid M, Girard O.Racket sports. In: McGarry T, O’Donoghue P, Sampaio J, eds. Handbook of Sports Performance Analysis. Routledge International Handbooks; 2015:404–414.
- 2.↑
Hulteen RM, Smith JJ, Morgan PJ, Barnett LM, Hallal PC. Global participation in sport and leisure-time physical activities: a systematic review and meta-analysis. Prev Med. 2017;95:14–25. doi:
- 3.↑
Pluim BM, Jansen MGT, Williamson S, et al. Physical demands of tennis across the different court surfaces, performance levels and sexes: a systematic review with meta-analysis. Sports Med. 2023;53(4):807–836. doi:
- 4.↑
Cádiz Gallardo MP, Pradas de la Fuente F, Moreno-Azze A, Carrasco Páez L. Physiological demands of racket sports: a systematic review. Front Psychol. 2023;14:295. doi:
- 5.↑
Méndez DH, Policastro PO, De Oliveira Silva D. Injury surveillance and training load methods used by health professionals in tennis: an online multinational survey. J Sport Rehabil. 2022;32(3):235–241. doi:
- 6.↑
Myers NL, Aguilar KV, Mexicano G, Farnsworth JL, Knudson D, Kibler W. The acute: chronic workload ratio is associated with injury in junior tennis players. Med Sci Sports Exerc. 2020;52(5):1196–1200. doi:
- 7.↑
Moreno-Pérez V, Prieto J, Del Coso J, et al. Association of acute and chronic workloads with injury risk in high-performance junior tennis players. Eur J Sport Sci. 2021;21(8):1215–1223. doi:
- 8.↑
Staunton CA, Abt G, Weaving D, Wundersitz DWT. Misuse of the term ‘load’ in sport and exercise science. J Sci Med Sport. 2022;25(5):439–444. doi:
- 9.↑
Impellizzeri FM, Rampinini E, Coutts AJ, Sassi A, Marcora SM. Use of RPE-based training load in soccer. Med Sci Sports Exerc. 2004;36(6):1042–1047. doi:
- 10.↑
Impellizzeri FM, Marcora SM, Coutts AJ. Internal and external training load: 15 years on. Int J Sports Physiol Perform. 2019;14(2):270–273. doi:
- 11.↑
Impellizzeri FM, Shrier I, Mclaren SJ, et al. Understanding training load as exposure and dose. Sports Med. 2023;53:1667–1679. doi:
- 12.↑
Keaney EM, Reid M. Quantifying hitting activity in tennis with racket sensors: new dawn or false dawn? Sports Biomech. 2018;19(6):831–839. doi:
- 13.↑
Buchheit M, Allen A, Poon TK, Modonutti M, Gregson W, Di Salvo V. Integrating different tracking systems in football: multiple camera semi-automatic system, local position measurement and GPS technologies. J Sports Sci. 2014;32(20):1844–1857. doi:
- 15.↑
Dennis R, Farhart R, Goumas C, Orchard J. Bowling workload and the risk of injury in elite cricket fast bowlers. J Sci Med Sport. 2003;6(3):359–367. doi:
- 16.↑
Myers NL, Sciascia AD, Kibler WB, Uhl TL. Volume-based interval training program for elite tennis players. Sports Health. 2016;8(6):536–540. doi:
- 17.↑
Whiteside D, Cant O, Connolly M, Reid M. Monitoring hitting load in tennis using inertial sensors and machine learning. Int J Sports Physiol Perform. 2017;12(9):1212–1217. doi:
- 18.↑
Kibler WB, Chandler TJ, Shapiro R, Conuel M. Muscle activation in coupled scapulohumeral motions in the high performance tennis serve. Br J Sports Med. 2007;41(11):745–749. doi:
- 19.↑
Girard O, Millet GP. Physical determinants of tennis performance in competitive tennis players. J Strength Cond Res. 2009;23(6):1867–1872. doi:
- 20.↑
Kovacs MS, Ellenbecker TS, Development P, et al. A performance evaluation of the tennis serve: implications for strength, speed, power, and flexibility training. Strength Cond J. 2011;33(4):22–30. doi:
- 21.↑
Baiget E, Corbi F, López J. Influence of anthropometric, ball impact and landing location parameters on serve velocity in elite tennis competition. Biol Sport. 2023;40(1):273–281. doi:
- 22.↑
Girard O, Micalleff JP, Millet GP. Influence of restricted knee motion during the flat first serve in tennis. J Strength Cond Res. 2007;21(3):950–957.
- 23.↑
Reid M, Elliott B, Alderson J. Lower-limb coordination and shoulder joint mechanics in the tennis serve. Med Sci Sports Exerc. 2008;40(2):308–315. doi:
- 24.↑
Baiget E, Colomar J, Corbi F. Upper-limb force-time characteristics determine serve velocity in competition tennis players. Int J Sports Physiol Perform. 2022;17(3):358–366. doi:
- 25.↑
Martin C, Bideau B, Ropars M, Delamarche P, Kulpa R. Upper limb joint kinetic analysis during tennis serve: assessment of competitive level on efficiency and injury risks. Scand J Med Sci Sport. 2014;24(4):700–707. doi:
- 26.↑
Martin C, Kulpa R, Ezanno F, Delamarche P, Bideau B. Influence of playing a prolonged tennis match on shoulder internal range of motion. Am J Sports Med. 2016;44(8):2147–2151. doi:
- 27.↑
Campbell A, Straker L, O’Sullivan P, Elliott B, Reid M. Lumbar loading in the elite adolescent tennis serve: link to low back pain. Med Sci Sports Exerc. 2013;45(8):1562–1568. doi:
- 28.↑
Colomar J, Peña J, Vicens-Bordas J, Baiget E. Minimizing maladaptations derived from tennis serve training and competition loads. Strength Cond J. 2023;10:824. doi:
- 29.↑
Marfarlane TUS, Tannath JS, Vincent GK. The validity and reliability of global positioning systems in team sport: a brief review. J Strength Cond Res. 2015;30(5):1470–1490.
- 30.
Alanen AM, Räisänen AM, Benson LC, Pasanen K. The use of inertial measurement units for analyzing change of direction movement in sports: a scoping review. Int J Sports Sci Coach. 2021;16(6):1332–1353. doi:
- 31.↑
Malone JJ, Lovell R, Varley MC, Coutts AJ. Unpacking the black box: applications and considerations for using GPS devices in sport. Int J Sports Physiol Perform. 2017;12(2):588–595. doi:
- 32.↑
Taylor JB, Wright AA, Dischiavi SL, Townsend MA, Marmon AR. Activity demands during multi-directional team sports: a systematic review. Sports Med. 2017;47(12):2533–2551. doi:
- 33.↑
Miguel M, Oliveira R, Loureiro N, García-Rubio J, Ibáñez SJ. Load measures in training/match monitoring in soccer: a systematic review. Int J Environ Res Public Health. 2021;18(5):2721–26. doi:
- 34.↑
Reid M, Morgan S, Whiteside D. Matchplay characteristics of Grand Slam tennis: implications for training and conditioning. J Sports Sci. 2016;34(19):1791–1798. doi:
- 35.↑
Whiteside D, Reid M. External match workloads during the first week of Australian open tennis competition. Int J Sports Physiol Perform. 2017;12(6):756–763. doi:
- 36.↑
Faude O, Meyer T, Rosenberger F, Fries M, Huber G, Kindermann W. Physiological characteristics of badminton match play. Eur J Appl Physiol. 2007;100(4):479–485. doi:
- 37.↑
Girard O, Chevalier R, Habrard M, Sciberras P, Hot P, Millet G. Game analysis and energy requirements of elite squash. J Strength Cond Res. 2007;21(3):909–914. doi:
- 38.↑
Mellado-Arbelo Ó, Baiget Vidal E, Usón MV. Análisis de las acciones de juego en pádel masculino profesional. Cult Cienc Deport. 2017;14(42):191–201.
- 39.↑
Reid M, Duffield R, Dawson B, Baker J, Crespo M. Quantification of the physiological and performance characteristics of on-court tennis drills. Br J Sports Med. 2008;42(2):146–151. doi:
- 40.↑
Rigozzi CJ, Vio GA, Poronnik P. Application of wearable technologies for player motion analysis in racket sports: a systematic review. Int J Sport Sci Coach. 2022;18:2321–2346.
- 41.↑
Tricco AC, Lillie E, Zarin W, et al. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann Intern Med. 2018;169(7):467–473. doi:
- 42.↑
Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:71. doi:
- 43.↑
Godin K, Stapleton J, Kirkpatrick SI, Hanning RM, Leatherdale ST. Applying systematic review search methods to the grey literature: a case study examining guidelines for school-based breakfast programs in Canada. Syst Rev. 2015;4:138. doi:
- 44.↑
Castillo AV. Review of ITF-approved Player Analysis Technology (PAT). ITF Coach Sport Sci Rev. 2023;(89):27–35. doi:
- 45.↑
Gogotak. Choreiking 2. Published 2021. Accessed February 28, 2023. https://gogotak.co.kr/bbs/content.php?co_id=CHOREIKING2
- 46.↑
Racketware. Racketware 1.0. Published 2021. Accessed February 28, 2023. https://www.racketware.co.uk/
- 47.↑
Sony. Sony Smart Sensor. Published 2014. Accessed February 28, 2023. https://www.smarttennissensor.sony.net
- 48.↑
Myers NL, Kibler WB, Axtell AH, Uhl TL. The Sony Smart Tennis Sensor accurately measures external workload in junior tennis players. Int J Sports Sci Coach. 2019;14(1):24–31. doi:
- 50.↑
Hadžić V, Germič A, Filipčič A. Validity and reliability of a novel monitoring sensor for the quantification of the hitting load in tennis. PLoS One. 2021;16:339. doi:
- 51.↑
Babolat. Babolat Play. Published 2013. Accessed February 28, 2023. https://help.babolatplay.com/hc/en-us/articles/360017139418-PLAY-POP-Discontinuation-of-connected-services
- 53.↑
Edelmann-Nusser A, Raschke A, Bentz A, Montenbruck S, Edelmann-Nusser J, Lames M. Validation of sensor-based game analysis tools in tennis. Int J Comput Sci Sport. 2019;18(2):49–59. doi:
- 54.↑
Lin J, Chang CW, Wang CH, et al. Design and implement a mobile badminton stroke classification system. In: 19th Asia-Pacific Network Operations and Management Symposium: Managing a World of Things, APNOMS 2017. 2017:235–238.
- 57.↑
Artengo. Artengo Personal Coach. Published 2013. Accessed February 28, 2023. https://www.artengo.fr
- 58.↑
Coollang. Coollang Sensor. Published 2020. Accessed February 28, 2023. https://www.coollang-global.com
- 59.↑
Babolat. Babolat Pop. Published 2015. Accessed February 28, 2023. https://help.babolatplay.com/hc/en-us/articles/360017139418-PLAY-POP-Discontinuation-of-connected-services
- 62.↑
Coollang. Smart Xiaoyu 2. Published 2015. Accessed February 28, 2023. http://www.coollang-global.com
- 63.↑
Coollang. Koospur Xiaoyu 3.0. Published 2017. Accessed February 28, 2023. http://www.coollang-global.com
- 64.↑
García GD, Coll JS, Infantes SC, et al. Validation of wearables for technical analysis of tennis players. Int J Racket Sports Sci. 2022;4(2):56–60.
- 66.↑
Perri T, Reid M, Murphy A, Howle K, Duffield R. Prototype machine learning algorithms from wearable technology to detect tennis stroke and movement actions. Sensors. 2022;22(22):868. doi:
- 67.↑
Troiano RP, McClain JJ, Brychta RJ, Chen KY. Evolution of accelerometer methods for physical activity research. Br J Sports Med. 2014;48(13):1019–1023. doi:
- 68.↑
Actofit. Actofit Badminton Tracker. Published 2019. Accessed February 28, 2023. https://actofit.com/
- 69.↑
Fitzpatrick A, Stone JA, Choppin S, Kelley J. Important performance characteristics in elite clay and grass court tennis match-play. Int J Perform Anal Sport. 2019;19(6):942–952. doi:
- 70.↑
Whiteside D, Elliott B, Lay B, Reid M. The effect of racquet swing weight on serve kinematics in elite adolescent female tennis players. J Sci Med Sport. 2014;17(1):124–128. doi:
- 71.↑
Bonato M, Maggioni MA, Rossi C, Rampichini S, La Torre A. Relationship between anthropometric or functional characteristics and maximal serve velocity in professional tennis players. J Sports Med Phys Fitness. 2015;55(10):1157–1165.
- 72.↑
Colomar J, Baiget E, Corbi F, Muñoz J. Acute effects of in-step and wrist weights on change of direction speed, accuracy and stroke velocity in junior tennis players. PLoS One. 2020;15(3):e0230631. doi:
- 74.↑
Gogotak. Choreiking 1. Published 2020. Accessed February 28, 2023. http://gogotak.co.kr/bbs/content.php?co_id=CHOREIKING1
- 75.↑
BlastMotion. Blast Swing Analyzer. Published 2014. Accessed February 28, 2023. https://store.blastmotion.com/store/products/baseball/
- 76.↑
BlastMotion. Blast Personal Swing Trainer. Published 2022. Accessed February 28, 2023. https://store.blastmotion.com/store/products/baseball-st/
- 77.↑
Arccos. Arccos Caddie Smart Sensor (Gen3+). Published 2022. Accessed February 28, 2023. https://eu.arccosgolf.com/products/arccos-caddie-smart-sensors
- 78.
Arccos. Arccos Caddie Smart Grip. Published 2022. Accessed February 28, 2023. https://eu.arccosgolf.com/products/arccos-caddie-smart-grips
- 79.↑
BlastMotion. Blast Swing & Stroke Analyzer. Published 2021. Accessed February 28, 2023. https://blastmotion.com/products/golf/
- 79.↑
Quattrium. FWD Powershot 2. Published 2014. Accessed February 28, 2023. https://store.roverwiz.com/products/fwd-powershot-2-ice-hockey-stick-performance-sensor
- 81.↑
Jiao L, Wu H, Bie R, Umek A, Kos A. Multi-sensor golf swing classification using deep CNN. Procedia Comput Sci. 2018;129:59–65. doi:
- 82.↑
Hardegger M, Ledergerber B, Mutter S, et al. Sensor technology for ice hockey and skating. 2015 IEEE 12th International Conference Wearable Implant Body Sensors Networks, BSN. 2015.
- 83.↑
Vleugels R, Van Herbruggen B, Fontaine J, De Poorter E. Ultra-wideband indoor positioning and IMU-based activity recognition for ice hockey analytics. Sensors. 2021;21(14):650. doi:
- 84.↑
Burland JP, Outerleys JB, Lattermann C, Davis IS. Reliability of wearable sensors to assess impact metrics during sport-specific tasks. J Sports Sci. 2021;39(4):406–411. doi:
- 85.
Roell M, Roecker K, Gehring D, Mahler H, Gollhofer A. Player monitoring in indoor team sports: concurrent validity of inertial measurement units to quantify average and peak acceleration values. Front Physiol. 2018;9:141. doi:
- 86.
Roell M, Mahler H, Lienhard J, Gehring D, Gollhofer A, Roecker K. Validation of wearable sensors during team sport-specific movements in indoor environments. Sensors. 2019;19(16):458. doi:
- 87.
MacDonald K, Bahr R, Baltich J, Whittaker JL, Meeuwisse WH. Validation of an inertial measurement unit for the measurement of jump count and height. Phys Ther Sport. 2017;25:15–19. doi:
- 88.↑
Armitage M, Beato M, Mcerlain-naylor SA. Inter-unit reliability of IMU Step metrics using IMeasureU Blue Trident inertial measurement units for running-based team sport tasks. J Sports Sci. 2021;39(13):1512–1518. doi:
