Autonomous detection system for non‐hard‐hat use at construction sites using sensor technology

Jung Hoon Kim, Byung Wan Jo, Jun Ho Jo, Yun Sung Lee, Do Keun Kim

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


In this study, we present a novel method of detecting hard hat use on construction sites using a modified version of an off‐the‐shelf wearable device. The data‐transmitting node of the device contained two sensors, a photoplethysmogram (PPG) and accelerometers (Acc), along with two modules, a global positioning system (GPS) and a low‐power wide‐area (LoRa) network module. All the components were embedded into a microcontroller unit (MCU) in addition to the power supply. The receiving node included a server that displayed the results via both the Internet of Things (IoT) and smartphones. The LoRa network connected two nodes so that it could function in larger areas such as construction sites at a relatively low cost. The proposed method analyzes the data from a PPG sensor located on the hard hat chin strap and automatically notifies a manager when a worker is not wearing the required hard hat at the site. In addition, by utilizing the PPG sensor data, a heart rate abnormality‐detecting feature was added based on an age‐adjusted maximum heart rate formula. In validation tests, various PPG sensor locations and shapes were studied, and the results demonstrated the smallest error in the circular shaped sensor located at the upper neck (0.56%). Finally, an IoT monitoring page was created to monitor heart rate abnormalities while identifying hard hat use violations via both PCs and smart phones.

Original languageEnglish
Article number1102
Pages (from-to)1-11
Number of pages11
JournalSustainability (Switzerland)
Issue number3
StatePublished - 2021 Feb 1


  • Construction
  • Detection
  • Hard hat
  • Safety
  • Sensor technology
  • Web monitoring


Dive into the research topics of 'Autonomous detection system for non‐hard‐hat use at construction sites using sensor technology'. Together they form a unique fingerprint.

Cite this