Comfort modelling is a critical scientific barrier to reaching better thermal satisfaction in buildings. It allows designers to combine different cooling systems better to target comfortable low-energy buildings in hot and tropical climates. Increasing computer performance offers new perspectives to use more refined thermophysiological models against traditional normative ones. Also, new types of coupled cooling alternatives arise and set a need for adequate comfort assessment models. The proposed article presents a methodology for better understanding human discomfort based on sensory response in hot conditions. It is an entry point to develop better and calibrate more generic bio-heat models for comfort prediction in the building industry. This study is part of a 48-month project called CoolDown funded by the French Nation Research Agency. It presents the first four months of field measurement. Preliminary results already give first insights into how relative humidity is predominant in hot climates when overreaching 68% and how temperature range is significant in occupant satisfaction when relative humidity is on the high side.