Introduction

Contemporary retail spaces are undergoing a profound transformation: from transactional spaces, they are becoming experiential environments, characterized by shifts in atmosphere, fluctuations in density, short and intense periods and an increasing heterogeneity of user profiles. The foundational work in this field has long addressed comfort using single-sensory models based on thermal, luminous, or acoustic stability. However, the analysis of the theoretical framework conducted in our research shows that these approaches are now insufficient to grasp the true nature of human experience in retail environments. Comfort is constructed at the intersection of multisensory perception, attention, emotions, movement intentions, and situated behaviors.

Our study contributes to this critique. It aims to develop an integrated model of multisensory comfort, capable of simultaneously articulating physical, perceptual, emotional and behavioral dimensions, while taking into account the temporal dynamics specific to retail spaces. It is based on a detailed analysis of the scientific barriers identified in the literature and explained in our manuscript, as well as on the need for a paradigm shift in the architectural design of commercial environments.

Scientific results and key challenges

A thorough analysis of the literature reveals six major challenges:

• The absence of an integrative model of multisensory comfort capable of simultaneously accounting for thermal, lighting, acoustic, olfactory, spatial and social dimensions. Unlike domestic or office spaces, commercial spaces present a high sensory density combined with highly variable exposure times.

• The lack of connection between objective measurements and subjective perceptions, often illustrated by the performance gap. In commercial contexts, this gap is amplified by rapid environmental transitions, locomotor activity, and the presence of other users.

• The failure to consider temporal dynamics, which are nevertheless essential for understanding how comfort changes during movement.

• The under-regarding of situated behaviors as indicators of comfort. Our analyses show that they nevertheless constitute robust markers: trajectories, speeds, stops, avoidance maneuvers.

• The difficulty lies in representing inter-individual heterogeneity: different sensitivities, motivations, levels of attention and needs.

• The lack of adaptive devices in commercial environments persists, despite the increasing availability of ambient technologies (dynamic lighting, thermal control, responsive facades, sound diffusion, etc.).

These limitations call for a reconfiguration of the core concept of comfort, understood not as a state, but as an adaptive, distributed, multi-sensory, and contextualized process.

Theoretical positions and foundations of the integrated model

The work of Gallissot [1], fundamental to this research, demonstrates that comfort should be considered a probabilistic, multisensory, and contextual phenomenon resulting from a non-linear integration of stimuli. This view is consistent with the framework of multisensory integration developed by Spence [2], where modalities constantly interact, producing effects of dominance, attenuation or reinforcement.

Similarly, the research of Ranne [3] and Altay [4] on atmospheres shows that sensory experience in semi-public environments depends more on perceptual transitions and dynamics than on the absolute values of stimuli. The notion of atmosphere is understood there as a perceptual operator, guiding attention, emotions and behavior.

Finally, Mélodie Sannier's [5] work on trajectory analysis using capacitive sensors and on behavioral recognition in environments provides a crucial methodological foundation for linking micro-variations in behavior and perceptual states.

These frameworks converge on the need for a truly multidimensional model, capable of simultaneously integrating physical stimuli, emotional parameters, adaptation strategies, situated behaviors, user profiles and temporal dimensions.

Proposed methodology

The methodology developed in our research is based on a hybrid approach, structured in three complementary parts:

1. Analytical part: formalizing the dynamics of comfort

Based on the manuscript, theoretical analysis and a review of the state of the art, we could identify the constituent dimensions of multisensory comfort, their interactions, the perceptual mechanisms involved and the behavioral variables correlated with states of comfort. This analysis leads to the formulation of a conceptual framework aimed at clarifying the links between perceptions, behaviors and environments.

2. Experimental part: immersive and in situ application

We are developing an immersive experience that allows the independent manipulation of light, acoustic, thermal and spatial parameters, combined with the analysis of situated behaviors (trajectories, speeds, stopping zones) and detailed subjective surveys based on perceived contextual variations. These experiences allow us to link: stimuli → perceptions → behaviors → interpretation of comfort

3. Computational component: towards a multi-agent model

To model the variability and dynamics of comfort, we construct a computational framework based on sensory, perceptual, behavioral, and adaptive agents. This model rests on the idea that agents constantly perceive, interpret, and adjust their behaviors according to detected environmental transitions.

Conclusion and perspectives 

Preliminary results show that situated behaviors are reliable indicators of comfort, that the dynamics of transitions strongly influence the experience, that normative conformity does not predict the actual experience, that users adopt rapid and contextual adaptation strategies and that the multi-agent approach offers a relevant framework for simulating this variability.

Future directions include the formalization of the integrated model, the integration of data from advanced sensory devices for continuous detection, the creation of a prototype adaptive retail space and on-site experimentation with real-time modulation scenarios.

References

[1] Gallissot, M. (2012). Modéliser le concept de confort dans un habitat intelligent : du multisensoriel au comportement.
[2] Spence, C. (2020). Sense of place: Architectural design for the multisensory mind.
[3] Ranne, J. (2019). Designing multisensory experiences in the built environment.
[4] Altay, B. (2021). Multisensory experience of public interiors.
[5] Sannier, M. et al. (2022). Extracting walking trajectories at home from a capacitive proximity sensing floor.