Cognitive Performance in Construction Safety Behavior Based on Physiological Measurement 


Student thesis: Doctoral Thesis

View graph of relations



Awarding Institution
Award date6 Sep 2022


The construction workplace poses enormous concerns and challenges to behavior and safety performance owing to its immense complexity, uncertainty, and organized nature over the past decades. Numerous researchers have been engaged in safety promotion topics, decrypting intertwined safety causes, inferring causality backward from accidents that have occurred, employing emerging sensors or wearable technologies for proactive monitoring and prevention alerts, or evolving novel applied management theories from the individual and organizational levels. Construction safety issues remain human-centered; thus, comprehending the interactions and process chains between the exterior physical world components and interior cognitive processing can facilitate the refinement of behaviors to minimize safety risks and avoid worse consequences. The complexity of the environment, the diversity of activities prevailing in the external world, and the differences in individual cognition have a substantial impact on the entire human chain of information sensation, perception, cognition processing, and behavior execution, as the interaction process involves multiple construction scenarios, multifarious activities stemming from different construction stages, and a wide variety of labor from diverse disciplines. Traditional research approaches are no longer adequate for deep cognitive investigation. Integrating physiological quantification and wearable devices enables it achievable and promising to explore the underlying physiological mechanisms of ongoing safety-critical scenarios from a cognitive perspective.

Given the basic points of departure, this study merged psychological experimental paradigms with cutting-edge electrophysiological technologies such as brain EEG topography and eye-based movement tracking technique, combining with the construction site safety scenes to scrutinize cognitive performance in the typical context of safety hazard identification, safety situation awareness for construction safety elements, dashboard monitoring of safety data. Depending on the environment's peculiarities, factors extracted from different human-centered scenarios are exploited. Specifically, the impact of environmental noise on cognitive performance was considered in the identification of fall hazards; and the impact of task-based workload on situational awareness was examined in the scene of safety-critical elements perception; the influence of individual cognitive style on the cognitive load of safety dashboard information perception was evaluated. Based on the exploration of the link between cognitive performance and safe behaviors, the feasibility of using wearable EEG devices to monitor cognitive performance for unsafe behavior avoidance was investigated. The primary findings revolved around the effects of environmental, task-type, and individual variables on cognitive performance as follows: 1) The effects of environmental noise on safety fall hazard identification illustrated that task-irrelevant noise contents and volume levels pose varying deleterious effects on attention, mental workload, and stress from cerebral metrics. Conversational noise and music rhythm, in particular, depressed attention levels substantially, whereas moderate volume noise undermined attention but considerably increased stress levels; 2): A two-factor (workload versus conditions) within-subject N-back paradigm design for safety-critical elements perception with EEG waveforms acquisition revealed that situation awareness would be depleted progressively as workload increases. The P300 component by ERPs at the occipital lobe, as well as band power localized in alpha and the beta spectral band, were recognized as beneficial biomarkers of the first stage of situation awareness; 3) In a dashboard interaction experiment, the individual difference in cognitive style plays an important role in perceived cognitive load measured by oculomotor measures. The cognitive load was lower in the field-independent (FI) group than in the field-dependent (FD) group reflected by fixation duration, fixation count, and saccade count. Although there was no statistical difference between the two groups in terms of dwell percentage in the functional area, relevant or irrelevant information area, the FD group had a shorter average fixation duration in the relevant information area, and the average fixation duration and the number of fixations in the irrelevant information area were longer in the FD group than in the FI group; 4) Evoked distractions were detected in a sustained attention response (SART) task containing graphic stimuli of hazardous opening and reveled the superior performance of frequency features on the temporal lobe, validating the feasibility of monitoring and managing cognitive performance.

Physiological measurements establish the bridge between cognitive status and behavioral shaping, while also revealing the cognitive performance of individuals to organize and process safety information. Exploring the impact of these variables on cognition is essential for the development of physiological measurement techniques to examine the causes of unsafe behaviors, as well as the feasibility studies of real-time perception and monitoring of human physiological cognitive states utilizing wearable sensors or devices. Safe cognitive state assessment based on acquired physiological data and the establishment of precise cerebral or physical features computational system to provide early warnings as well as to forecast probable aberrant states would be beneficial and desirable in the future.

    Research areas

  • Cognitive performance, Physiological measurement, Construction safety, Electroencephalography (EEG)