Publications / 2024 Proceedings of the 41st ISARC, Lille, France
Ensuring workers safety at construction sites is complicated as protective measures often involve the tasks of planning, monitoring, and mitigation at the same time. Despite traditional methods during the pre-construction and construction phases that
require time-consuming and manual efforts, poor risk assessment and situational awareness can easily lead to unplanned mishaps in detecting and eliminating risk. Semi-automated rule-based risk assessment approaches as they predominantly exist in research (ref. SafeConAI) are capable of designing out known hazards before they appear in the workplace. These, however, tend not to be interoperable with other emerging technology tasked to monitor how well safety is practiced on the construction site. This paper presents a method for enhanced safety incident analysis by fusing preidentified hazard zones that remain in construction schedules (after SafeConAI has been applied to a 4D BIM) with high-precision trajectory data (using RTK-GNSS) of pedestrian workers and heavy construction equipment. A reallife case study validates the methods feasibility yielding, aside from basic statistical spatiotemporal counting of incident numbers and precise locations between the pedestrian workforce and construction equipment, also new insights into the right size of the so-defined protective safety envelopes that should surround the construction machinery. These promising results still require further investigation into the practical applicability, for example, testing the effectiveness of sharing the detailed personalized feedback that becomes now available.