Performance Feedback for Cable Machine Strength Exercises Using Smartphone’s Inertial Data
Name
Robert Altmäe
Abstract
For the purpose of endurance training, devices have been produced for many years that allow measuring pace, heart rate, step length, distance covered, and thanks to this find the amount of spent energy and other performance parameters. For this, integrated inertial measurment units, or IMUs, are used. However, for strength training, such IMU-based solutions are not as common. For performance feedback, manual measurement of repetitions and used weights has to be done. Even then, motion is still not measured. To solve this, a method is proposed in this thesis to fetch useful exercise feedback metrics from acceleration data.
These metrics are:
- energy consumption in kilocalories
- repetition count
- peak power in Newtons.
This work describes how the movement of a cable machine’s weight stack in one axis was measured with an acceleration sensor of a mobile phone, and how the data was processed and useful indicators were found for exercise feedback. An overview of the strength exercise performed for data collection was given, the data collection hardware and software were described, and the results were analyzed and visualized. A baseline comparison was done to compare the method’s results with theoretical calculations, showing promising results. In addition, problem areas related to noise and measuring were pointed out and possible ideas for further development are proposed. As a secondary part of the thesis, a software solution was developed in the form of a user interface and a server, the technological choices and implementation of which are described. A link to the source code repository is also given. This research has the potential to be used in health applications on smartphones.
These metrics are:
- energy consumption in kilocalories
- repetition count
- peak power in Newtons.
This work describes how the movement of a cable machine’s weight stack in one axis was measured with an acceleration sensor of a mobile phone, and how the data was processed and useful indicators were found for exercise feedback. An overview of the strength exercise performed for data collection was given, the data collection hardware and software were described, and the results were analyzed and visualized. A baseline comparison was done to compare the method’s results with theoretical calculations, showing promising results. In addition, problem areas related to noise and measuring were pointed out and possible ideas for further development are proposed. As a secondary part of the thesis, a software solution was developed in the form of a user interface and a server, the technological choices and implementation of which are described. A link to the source code repository is also given. This research has the potential to be used in health applications on smartphones.
Graduation Thesis language
English
Graduation Thesis type
Master - Data Science
Supervisor(s)
Naveed Muhammad
Defence year
2024