Development and Open-Source Implementation of JOS-3: A Human Thermoregulation Model for Diverse Environmental Applications

Accurate evaluation of human thermal comfort is essential for designing thermally acceptable environments. While widely used indices such as Predicted Mean Vote (PMV) and Standard New Effective Temperature (SET*) perform well under uniform and steady-state conditions, their applicability is limited in non-uniform or transient environments. In response, we developed a new open-source human thermoregulation model, JOS-3, to predict physiological responses more accurately under diverse environmental conditions. These include the representation of brown adipose tissue (BAT) activity under cold acclimation, aging effects, and solar heat gain at the skin surface due to shortwave radiation. In addition, we revised the algorithms for shivering thermogenesis, sweat distribution, and basal metabolic rate estimation. The model consists of 83 thermal nodes and employs the backward difference method for solving body temperature and physiological responses over time. JOS-3 was implemented in Python 3 and is publicly available under GitHub (https://github.com/TanabeLab/JOS-3), facilitating collaboration and use across research institutions. The present study aims to further enhance JOS-3 by addressing feedback from international projects and domestic users, such as emergency medical professionals requiring deep body temperature predictions. By expanding its applicability and accuracy, JOS-3 contributes not only to building environmental research but also to interdisciplinary fields involving human thermal physiology.