Dynamic Adaptation in Thermal Comfort Modeling

16 concepts10 papers1 proposal

Traditional heating, ventilation, and air conditioning (HVAC) systems operate on a simple premise: if you maintain indoor air temperature within a fixed, standardized range, occupants will be comfortable. This approach relies on static, physics-based models developed under steady-state laboratory conditions.

In the real world, however, this assumption frequently breaks down. Automated systems often overheat or undercool spaces, leading to both energy waste and occupant dissatisfaction. The fundamental flaw of static models is that they treat humans as passive thermometers. In reality, building occupants are active, adaptive agents who constantly adjust their behavior, clothing, and expectations to achieve thermal satisfaction.

To bridge this gap, a growing body of research focuses on integrating human adaptation into thermal comfort modeling. By understanding why and how humans adapt, engineers and computer scientists can design climate control systems that are both more energy-efficient and closely aligned with actual human sensations.

The Failure of Static Baselines

Standardized predictive comfort models—such as the widely used ASHRAE 55 standard—frequently fail in dynamic, real-world environments. As Kumar et al. observe in their work on smart comfort sensing, these universal metrics create significant discrepancies between calculated comfort indices and actual subjective human sensations.

This failure occurs because static models ignore the active ways individuals compensate for environmental conditions. When Purdon et al. evaluated traditional HVAC control in Model-free HVAC control using occupant feedback, they noted that complex thermal models struggle precisely because they cannot capture subjective human adaptability. Similarly, Zhu et al. point out in their study of vernacular dwellings that Western-centric standards fail to capture the unique adaptive behaviors and preferences of local populations, particularly when dynamic factors like manual window operation are left unmodeled.

Human thermal adaptation operates across three primary dimensions: behavioral, psychological, and physiological.

Behavioral Adaptation: The Active Occupant

Behavioral adaptation is the most immediate way humans regulate their thermal state. If a room is too cold, an occupant might put on a jacket or close a window; if it is too warm, they might adjust a fan.

When environmental models omit these actions, their predictive power degrades. Gao et al. emphasize this in their transfer learning framework, noting that personal attire and window adjustments serve as critical mediators of thermal satisfaction. Without integrating this behavioral data, systems cannot accurately predict whether an occupant is comfortable.

[Environmental Change] ──> [Behavioral Adjustment] ──> [Altered Thermal Sensation]
         │                                                      │
         └─────────── (Ignored by Static Models) ───────────────┘

This effect is vividly demonstrated in regions without centralized climate control. Li et al. investigated winter thermal comfort in China's Hot Summer and Cold Winter climate zone in their field study. They found that residents routinely tolerate indoor temperatures well below standard hygiene limits by wearing significantly more clothing than standard models assume. This high level of personal insulation serves as a primary adaptive mechanism. By modifying standard predictive models with adaptive thermal indices that account for these clothing adjustments, the authors achieved a far more accurate assessment of real-world comfort.

The importance of personal factors extends to automated office environments as well. Κόντες et al. demonstrated through building simulations in their research that relying solely on dry-bulb air temperature setpoints is insufficient. To prevent discomfort, control loops must dynamically adapt to personal variables like clothing insulation alongside environmental factors like humidity. Even in extreme public health contexts, such as modeling heatstroke risk, Deng et al. note in their meteorological analysis that omitting spontaneous behavioral adaptations—such as choosing to remain indoors during peak heat—makes it difficult to determine whether health outcomes are driven purely by weather thresholds or by human action.

Psychological Adaptation: Shifting Seasonal Baselines

Thermal comfort is not just a matter of skin temperature; it is heavily influenced by expectations. A room temperature of 24°C might feel uncomfortably warm in the depths of winter but refreshingly cool during a summer heatwave.

Static environmental models overlook this psychological adaptation, causing automated systems to overheat or undercool spaces unnecessarily. To address this, Lim et al. proposed an HVAC-aware occupancy scheduling framework that replaces fixed temperature bounds with dynamic comfort models, allowing ideal setpoints to shift based on fluctuating outdoor conditions.

Capturing these shifting expectations requires modeling the temporal context of the occupant. In A Personalised Thermal Comfort Model Using a Bayesian Network, Auffenberg et al. explicitly incorporated seasonal variables into a probabilistic framework. This mechanism recognizes that a user's thermal neutrality point is a moving target. By tracking the time of year and outdoor conditions, their model shifts the baseline comfort expectation over time, successfully preventing the unnecessary energy expenditures common to "smart" thermostats that rely on static preferences.

Perceptual Drift in Steady-State Environments

Even when an HVAC system successfully maintains a perfectly constant temperature within a theoretically "acceptable" range, human satisfaction is not guaranteed. Human perception naturally drifts over time, even in steady-state environments.

An empirical study on cooling regulation characteristics documented this phenomenon. When residents were placed in steady-state conditions between 25°C and 27°C, they experienced a gradual drift toward discomfort. Rather than remaining static, occupants engaged in a predictable sequence of micro-adaptations:

  1. Micro-behavioral defense: Occupants initially used head and face movements to alleviate subtle discomfort.
  2. Manual intervention: If the drift continued, occupants manually adjusted the thermostat—typically requesting a 1°C increase—every 20 to 30 minutes.

This pattern proves that human comfort cannot be solved as a static optimization problem; it requires continuous, dynamic feedback.

Designing Human-Centric Climate Control

Understanding the mechanisms of human adaptation allows researchers to design control systems that trade rigid environmental constraints for massive energy savings.

If a system knows that occupants will naturally adjust their clothing or tolerate higher temperatures due to seasonal expectations, it can allow the indoor climate to "drift" closer to outdoor ambient levels. Purdon et al. implemented this exact strategy by bypassing complex thermal modeling entirely in favor of direct occupant feedback via smartphone voting. By letting the temperature drift dynamically based on real-time human comfort, their system achieved energy savings of up to 60% with negligible impact on user satisfaction.

Ultimately, the shift from static physics-based equations to dynamic, adaptive comfort models represents a deeper understanding of the built environment. By treating the human occupant as an active participant rather than a fixed variable, modern climate control systems can achieve true thermal harmony while drastically reducing building energy footprints.

Key Foundational Work

An Energy-Efficient Smart Comfort Sensing System Based on the IEEE 1451 Standard for Green Buildings (2014) — contributed 1 concept elements to this direction.
Model-free HVAC control using occupant feedback (2013) — contributed 1 concept elements to this direction.
History & citationVersion 1 · 2026-10-04

Current · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_clothing_based_thermal_adaptation_29794070,
  title = {clothing-based thermal adaptation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/clothing-based-thermal-adaptation/rdr_29794070eae24a99af495db1e78a5c26},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_clothing_based_thermal_adaptation_29794070_v1,
  title = {clothing-based thermal adaptation},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/clothing-based-thermal-adaptation/rdr_29794070eae24a99af495db1e78a5c26/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}
About

These directions overlap with this one through shared papers.

See something wrong? Report a mistake