Abstract: Systems, methods, and computer program products for human-centered safety using physiological indicators to identify and/or predict potential hazards are disclosed. An example method includes receiving physiological data associated with a plurality of physiological parameters of a worker from a wearable device of the worker. The physiological data may be monitored to detect at least one potential hazard based on a hazard classifier. The hazard classifier may include at least one machine learning model trained based on historical physiological data associated with the plurality of physiological parameters. At least one communication may be communicated based on the potential hazard(s) and/or the physiological data.