Publications / 2016 Proceedings of the 33rd ISARC, Auburn, USA

An Integrated INS-GPS-Raspberry Pi System Using the Time-Sphere Model for Real-Time Identification of Struck-by-Equipment Hazard

Jun Wang, Shuming Du and Saiedeh Razavi
Pages 1032-1040 (2016 Proceedings of the 33rd ISARC, Auburn, USA, ISBN 978-1-5108-2992-3, ISSN 2413-5844)
Abstract:

Struck-by-equipment hazard, i.e, contact collision between worker-on-foot and construction equipment or between equipment and equipment, is a major cause of injuries and fatalities in construction. A variety of smart technologies and systems have been developed and utilized to alleviate the risks of this type of hazard. However, most of the existing technologies and systems perform with a high rate of false alarms which have critically impeded their implementations on real construction sites. Therefore, a system performing more accurately and in real-time needs to be developed to enhance construction safety. In this context, a real-time system integrating with the 4D Time-Sphere model is developed to timely identify and alarm struck-by-equipment hazards with reduced false alarms. This system is comprised of Global Positioning System (GPS), Inertial Navigation System (INS), Raspberry Pi micro-processor and the developed Time-Sphere model. Corresponding actuation will be triggered if an unsafe situation is identified, i.e., struck-by-equipment hazard(s) are or to be presented. A controlled field experiment was conducted to evaluate the feasibility of the integrated system and the effectiveness of the Time-Sphere model in reducing false alarms. The obtained results show that average 53% of the alarms generated by the prevalent method can be reduced by the Time-Sphere model. All of the reduced alarms are false positives. The average false positive rate of the prevalent method is 29%. Moreover, it is demonstrated that the developed system is feasible and effective in identifying struck-by-equipment hazards in real time and potentially enhancing construction safety.

Keywords: Real-Time Hazard Identification, Construction Safety, Technological Innovation, Raspberry Pi, Struck by, False Alarm, 4D Model, Inertial Navigation System