Performance Management for Athletes via Artificial Intelligence Using Sensor Data

Yasin Zencir; Emrah Aydemir

Article Details

Performance Management for Athletes via Artificial Intelligence Using Sensor Data

Authors:

Yasin Zencir | Emrah Aydemir

Publisher:

Zonguldak Bulent Ecevit University

Main Topics:

Bilgisayar Yazılımı | Devreler ve Sistemler

Journal:

Karaelmas Fen ve Mühendislik Dergisi Vol.15 Issue 1, pp.56-74

Document Type:

Article

Publication Date:

4/2025

Document Language:

Turkish

Document ID (Org No.):

2001345564

pISSN:

2146-7277

Keywords:

Abstract:

 Today, athletes performance data is obtained through various sensor-based devices and later used as a performance evaluation tool through different methods. In this study, performance data of athletes was collected using a device designed with an Arduino microcontroller board, obstacle sensors, and nRF wireless communication technology. This data was analyzed using supervised machine learning techniques to develop a performance evaluation model. The model provides a scientific, AI-based performance management system for both athletes and their coaches. The device was tested on 18 students aged 11-12, with each student completing 20 trials. Each student interacted with a total of 25 devices in a sequence, resulting in 500 data points per student. A total of 9,000 data points were collected, including the students weight, height, gender, and expert-provided evaluation labels, which were recorded in a database. The data was analyzed using Weka software with 32 different classification methods, with the highest accuracy rate of 95.8333 achieved using the MultilayerPerceptron method. The device and developed models can be applied not only in the sports field but also in healthcare and education. For example, it can be used in hospitals to make physical therapy exercises more engaging. It can also be utilized in schools to enhance physical education classes and prepare students for sports competitions. Additionally, it can serve as a play material in early childhood education to improve childrens psychomotor skills.

 Today, athletes performance data is obtained through various sensor-based devices and later used as a performance evaluation tool through different methods. In this study, performance data of athletes was collected using a device designed with an Arduino microcontroller board, obstacle sensors, and nRF wireless communication technology. This data was analyzed using supervised machine learning techniques to develop a performance evaluation model. The model provides a scientific, AI-based performance management system for both athletes and their coaches. The device was tested on 18 students aged 11-12, with each student completing 20 trials. Each student interacted with a total of 25 devices in a sequence, resulting in 500 data points per student. A total of 9,000 data points were collected, including the students weight, height, gender, and expert-provided evaluation labels, which were recorded in a database. The data was analyzed using Weka software with 32 different classification methods, with the highest accuracy rate of 95.8333 achieved using the MultilayerPerceptron method. The device and developed models can be applied not only in the sports field but also in healthcare and education. For example, it can be used in hospitals to make physical therapy exercises more engaging. It can also be utilized in schools to enhance physical education classes and prepare students for sports competitions. Additionally, it can serve as a play material in early childhood education to improve childrens psychomotor skills.

 Today, athletes performance data is obtained through various sensor-based devices and later used as a performance evaluation tool through different methods. In this study, performance data of athletes was collected using a device designed with an Arduino microcontroller board, obstacle sensors, and nRF wireless communication technology. This data was analyzed using supervised machine learning techniques to develop a performance evaluation model. The model provides a scientific, AI-based performance management system for both athletes and their coaches. The device was tested on 18 students aged 11-12, with each student completing 20 trials. Each student interacted with a total of 25 devices in a sequence, resulting in 500 data points per student. A total of 9,000 data points were collected, including the students weight, height, gender, and expert-provided evaluation labels, which were recorded in a database. The data was analyzed using Weka software with 32 different classification methods, with the highest accuracy rate of 95.8333 achieved using the MultilayerPerceptron method. The device and developed models can be applied not only in the sports field but also in healthcare and education. For example, it can be used in hospitals to make physical therapy exercises more engaging. It can also be utilized in schools to enhance physical education classes and prepare students for sports competitions. Additionally, it can serve as a play material in early childhood education to improve childrens psychomotor skills.

 Günümüzde sensör tabanlı birçok cihaz yardımıyla sporcuların performans verileri elde edilmekte ve bu veriler daha sonra farkı yöntemler kullanılarak performans değerlendirme aracı olarak kullanılmaktadır. Bu çalışmada arduino mikrodenetleyici kartı, engel sensörü ve nRF kablosuz iletişim teknolojisi kullanılarak tasarlanan bir cihaz yardımıyla sporcular ile ilgili performans veriler elde edilmiştir. Bu veriler denetimli makine öğrenmesi yöntemi ile analiz edilerek bir performans belirleme modeli oluşturulmuştur. Bu model, sporcu ve sporcu çalıştırıcıları için bilimsel veriye dayalı ve yapay zeka tabanlı bir performans yönetim sistemi sağlamaktadır. Tasarlanan cihaz 11-12 yaş aralığındaki 18 öğrenci üzerinde uygulanmış ve her öğrencinin 20 deneme yapması sağlanmıştır. 5 cihaz arasında koşarak toplamda 25 cihaza uğramaları sağlanan her bir öğrenci için 500 veri elde edilmiştir. 18 öğrenciden elde edilen 9000 veri, öğrencilere ait kilo, boy, cinsiyet ve bir uzman tarafından verilen değerlendirme etiketi ile birlikte bir veri tabanına kaydedilmiştir. Bu veriler Weka programı ile 32 adet sınıflandırma yöntemi kullanılarak analiz edilmiş ve en başarılı oran 95.8333 ile MultilayerPerceptron yöntemi ile elde edilmiştir. Bu çalışmada tasarlanan cihaz ve oluşturulan modeller sadece spor alanında değil, sağlık ve eğitim alanlarında da kullanılabilir. Örneğin; hastanelerde fizik tedavi çalışmalarını eğlenceli hale getirmek için kullanılabilir. Okullarda beden eğitimi derslerini sevdirmek ve öğrencileri spor müsabakalarına hazırlamak için kullanılabilir. Ayrıca okul öncesi eğitiminde de hem oyun materyali olarak hem de çocukların psiko-motor becerilerinin geliştirilmesi amacıyla kullanılabilir.

Article Details

Performance Management for Athletes via Artificial Intelligence Using Sensor Data

Authors:

Yasin Zencir | Emrah Aydemir

Publisher:

Zonguldak Bulent Ecevit University

Main Topics:

Bilgisayar Yazılımı | Devreler ve Sistemler

Journal:

Karaelmas Fen ve Mühendislik Dergisi Vol.15 Issue 1, pp.56-74

Document Type:

Article

Publication Date:

4/2025

Document Language:

Turkish

Document ID (Org No.):

2001345564

pISSN:

2146-7277

Aerobic and anaerobic skills

model development

psychomotor skills

reactive agility.Aerobik ve anaerobik beceri

model oluşturma

psiko-motor beceri

reaktif çeviklik.

Issues

Article Details

Authors: Yasin Zencir | Emrah Aydemir

Publisher: Zonguldak Bulent Ecevit University

Main Topics: Bilgisayar Yazılımı | Devreler ve Sistemler

Journal: Karaelmas Fen ve Mühendislik Dergisi Vol.15 Issue 1, pp.56-74

Document Type: Article

Publication Date: 4/2025

Document Language: Turkish

Document ID (Org No.): 2001345564

pISSN: 2146-7277

Issues

Authors:

Yasin Zencir | Emrah Aydemir

Publisher:

Zonguldak Bulent Ecevit University

Main Topics:

Bilgisayar Yazılımı | Devreler ve Sistemler

Journal:

Karaelmas Fen ve Mühendislik Dergisi Vol.15 Issue 1, pp.56-74

Document Type:

Article

Publication Date:

4/2025

Document Language:

Turkish

Document ID (Org No.):

2001345564

pISSN:

2146-7277