Conference Agenda

A growing body of research has demonstrated that access to fresh air indoors is closely linked to improved cognitive function, reduced absenteeism, and greater overall workplace satisfaction. Yet, the consequences of inadequate ventilation are often difficult to isolate in occupied office settings. This paper draws on a unique, naturally occurring experiment in a high-performing Atlanta office building, where a mechanical failure in the dedicated outdoor air system (DOAS) resulted in a complete loss of mechanically ventilated air for two months during the winter of 2024.

Continuous environmental monitoring over the course of the year revealed that, during the HVAC malfunction period, indoor carbon dioxide (CO₂) concentrations routinely exceeded the ASHRAE recommended threshold of 1000 ppm and, on several occasions, rose above 2000 ppm during working hours. These values contrast with an average of 557 ppm when the system was operating normally, highlighting the magnitude of the deviation.

To explore the broader implications, air quality data from the malfunction period is compared alongside anonymized financial productivity metrics provided by the firm, as well as occupant satisfaction surveys. Preliminary survey results suggest a marked decline in perceived air quality, comfort, and satisfaction during the malfunction period, reinforcing the critical role of ventilation in supporting a healthy and effective workplace.

This case study offers compelling evidence of the tight coupling between ventilation, occupant experience, and workplace performance. It highlights the vulnerability of even high-performing buildings to system failures and the need for continuous monitoring and resilient HVAC design. These findings underscore the importance of treating air quality not only as a health factor but also as a vital asset for organizational productivity and wellbeing.

Accurate prediction of overall thermal comfort (OTC) in radiant heating systems is essential for creating energy-efficient and user-centered environments. Radiant heating systems have emerged as a promising heating system that can provide uniform thermal comfort for building occupants by reducing temperature asymmetries in indoor environments. This study investigates the application of machine learning (ML) to predict OTC using experimental data on local body parts' thermal comfort. Measurements of local body parts’ skin temperatures and subjective comfort votes were collected using radiant heating devices. Traditional indices, such as Predicted Mean Vote (PMV), often fail to account for localized discomfort caused by non-uniform radiant heating. This work builds on prior chamber experiments to present an advanced machine-learning (ML) framework for predicting overall thermal sensation (o_sensation) and overall thermal comfort (o_comfort) from local physiological and perceptual inputs collected under directed radiant heating. Leveraging an expanded dataset of 100,000 samples produced from the KIT LOBSTER chamber experiments (original pilot ≈12,000 samples), we evaluated seven ML algorithms—Random Forest, XGBoost, LightGBM, CatBoost, Gradient Boosting, Ridge Regression, and K-Nearest Neighbors—within a multi-output architecture. Models were assessed with MAE, RMSE, and R² and interpreted through node–performance maps, normalized radar charts, network graphs, histograms, boxplots, and feature-importance analysis. Tree-based ensemble models, particularly Random Forest and XGBoost, consistently outperformed linear and instance-based methods, showing tighter error distributions and higher explanatory power. o_sensation proved more tightly coupled to skin temperature signals than o_comfort, which exhibited higher variance likely due to cognitive and contextual influences. We discuss practical implications for adaptive radiant HVAC control, limitations due to chamber conditions and data modalities, and future directions including multimodal sensing, temporal modeling, and validation. The integration of physiological and environmental parameters offers a personalized approach, bridging gaps in traditional methods.

Thermal alliesthesia is a psycho-physiological mechanism that governs the conditions of thermal pleasure in architectural spaces. It includes thermal pleasure arising from physiological reactions in non-steady and non-uniform environments, defined as temporal and spatial alliesthesia, respectively. Recent studies have outlined that this concept can be expanded to include non-sensory and broader cognitive processes, such as acquired pleasure from previous memories and architectural experiences, acknowledging both the psychological and physiological impacts of thermal stimuli. It is essential to understand the parameters that impact occupants’ perception of thermal alliesthesia to develop a comprehensive approach to designing built environments that not only maintain thermal comfort but also promote thermal pleasure and occupants’ well-being. However, the parameters governing thermal alliesthesia in architectural settings remain unclear. In addition, an integrated framework that incorporates both the physiological and psychological dimensions of this mechanism is still lacking. This gap in the literature limits opportunities to design spaces that are both thermally engaging and comfortable to occupy.

This paper reports on a systematic literature review to deduce a comprehensive framework illustrating the mechanisms of the three aspects of thermal alliesthesia: temporal, spatial, and phenomenological. A bibliometric analysis conducted using CiteSpace examines the evolution of research, disciplinary fields, and emerging fronts. This is followed by a theoretical development of the parameters that govern thermal alliesthesia in historic buildings, as an example for applying the alliesthesia mechanism in architecture. This application focuses on identifying the architectural parameters of thermal alliesthesia associated with both the physiological dimension of temporal and spatial alliesthesia and the psychological dimension of phenomenological alliesthesia. The paper concludes by identifying the gaps in the literature that warrant further investigations of thermal alliesthesia in architecture. In addition, it develops a comprehensive framework that details the mechanisms and architectural parameters for managing thermal comfort and thermal pleasure in architecture.