Ravanelli Nicholas, Lefebvre KarLee, Brough Amy, Paquette Simon, Lin Wei
Heat Resilience & Performance Centre, Yong Loo Lin School of Medicine, National University of Singapore, 27 Medical Drive, #03-01, Singapore 117510, Singapore.
Department of Physiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 117593, Singapore.
Sensors (Basel). 2025 May 6;25(9):2926. doi: 10.3390/s25092926.
Consumer-grade wrist-based wearable devices have grown in popularity among researchers to continuously collect metrics such as physical activity and heart rate. However, manufacturers rarely disclose the preprocessing sensor data algorithms, and user-generated data are typically shared leading to data governance issues. Open-source technology may address these limitations. This study evaluates the validity of the Bangle.js2 for step counting and heart rate during lab-based validation and agreement with other wearable devices (steps: Fitbit Charge 5; heart rate: Polar H10) in free-living conditions. A custom open-source application was developed to capture the sensor data from the Bangle.js2. Participants ( = 47; 25 males; 27 ± 11 years) were asked to complete a lab-based treadmill validation (3 min stages at 2, 3, 4, and 5 mph) and stair climbing procedure followed by a 24 h free-living period. The Bangle.js2 demonstrated systematic undercounting of steps at slower walking speeds with acceptable error achieved at 5 km/h. During free-living conditions, the Bangle.js2 demonstrated strong agreement with the Fitbit Charge 5 for per-minute step counting (CCC = 0.90) and total steps over 24 h (CCC = 0.96). Additionally, the Bangle.js2 demonstrated strong agreement with the Polar H10 for minute-averaged heart rate (CCC = 0.78). In conclusion, the Bangle.js2 is a valid open-source hardware and software solution for researchers interested in step counting and heart rate monitoring in free-living conditions.
消费级腕戴式可穿戴设备在研究人员中越来越受欢迎,可用于持续收集身体活动和心率等指标。然而,制造商很少披露预处理传感器数据的算法,而且用户生成的数据通常会被共享,从而导致数据治理问题。开源技术或许可以解决这些限制。本研究评估了Bangle.js2在实验室验证期间进行步数计数和心率监测的有效性,以及在自由生活条件下与其他可穿戴设备(步数:Fitbit Charge 5;心率:Polar H10)的一致性。开发了一个定制的开源应用程序来捕获Bangle.js2的传感器数据。参与者(n = 47;25名男性;27±11岁)被要求完成基于实验室的跑步机验证(在2、3、4和5英里/小时的速度下进行3分钟阶段)和爬楼梯程序,随后是24小时的自由生活期。Bangle.js2在较慢步行速度下显示出步数计数存在系统性少计的情况,在5公里/小时时误差可接受。在自由生活条件下,Bangle.js2在每分钟步数计数(CCC = 0.90)和24小时总步数(CCC = 0.96)方面与Fitbit Charge 5显示出高度一致性。此外,Bangle.js2在平均每分钟心率方面与Polar H10显示出高度一致性(CCC = 0.78)。总之,对于有兴趣在自由生活条件下进行步数计数和心率监测的研究人员来说,Bangle.js2是一种有效的开源硬件和软件解决方案。