Towards human-centric construction automation: an SA-STAMP framework for safe human–autonomy teaming

Conventional Job Hazard Analysis (JHA) and current automation studies often fail to address the socio-technical complexities of the worksite, frequently dismissing accidents as isolated ‘human error’. To bridge this gap, this paper presents a context-aware automation design framework that combines Situation Awareness (SA) with the System-Theoretic Accident Model and Processes (STAMP) to safely facilitate Human–Autonomy Teaming (HAT) in heavy modular lifting.

Going beyond existing Systems-Theoretic Process Analysis (STPA) applications, the proposed SA-STAMP framework directly integrates real-world site conditions into the system's control structure. Rather than treating these factors as external operator failures, the framework addresses them as systemic flaws requiring robust feedback loops, defined authority boundaries, and intuitive interface designs. By condensing a detailed modular lifting case study, the paper demonstrates how adaptive SA mechanisms can enforce machine-readable safety constraints during lift planning and execution.