Development of a wearable activity tracker based on BBC micro:bit and its performance analysis for detecting bachata...

Area of Science:

• Wearable Technology
• Human-Computer Interaction
• Dance Analytics

Background:

• Wearable activity trackers are increasingly popular for data collection and analysis.
• Applications of wearable sensor data span various fields, including fitness and performance monitoring.
• Automated analysis of physical activities like dance can benefit from wearable technology.

Purpose of the Study:

• To develop a wearable activity tracking system for identifying basic bachata dance steps.
• To utilize the BBC micro:bit development board for real-time data acquisition and transmission.
• To create a user interface for automatic dance step detection and analysis.

Main Methods:

• A pair of smart ankle bracelets was created using the BBC micro:bit, featuring an accelerometer and Bluetooth.
• A dataset of six bachata dance steps was collected from ten participants.
• Squared Euclidean distance was used to analyze accelerometer data for step recognition.
• A Python and Tkinter-based user interface was developed for the detection system.

Main Results:

• The developed system demonstrated an accuracy rate of 79.2% in detecting bachata dance steps.
• Accelerometer data from ankle movements was effectively utilized for dance step identification.
• The system successfully processed synchronized dance steps across multiple participants.

Conclusions:

• The BBC micro:bit-based wearable system is a viable tool for automatic bachata dance step detection.
• The use of accelerometer data and Euclidean distance metric shows promise for dance analytics.
• Further development could enhance accuracy and expand the range of detectable dance moves.