Making Thermostats Great Again: Investigating Touch-Based Thermal Sensation for Manual Control of Thermostats

Yidong Huang, Gaëlle Calvary, Yann LaurillauAVI 2024 · DOI

TL;DR

This paper introduces a novel approach to thermostat control by integrating touch-based thermal feedback, aiming to bridge the gap between user actions and delayed environmental responses. By providing immediate tactile thermal cues, the system allows users to intuitively adjust temperature setpoints and anticipate comfort levels, leading to more efficient and satisfying control.

What Problem Does This Paper Address?

Modern thermostats, while offering advanced features, often suffer from poor usability and a lack of immediate feedback, leading to wasted energy and user frustration. The core issue is the "delayed feedback" problem: users adjust a thermostat, but the actual change in room temperature and resulting thermal comfort takes time to manifest. This disconnect makes it difficult for users to form an accurate mental model of how the system works and to effectively control it. Consequently, users often misuse thermostats, overriding automated settings and negating potential energy savings. This paper specifically addresses the challenge of improving manual thermostat control by investigating the potential of touch-based thermal sensation as an intuitive and immediate feedback mechanism.

Methods

The researchers developed a proof-of-concept (PoC) thermostat with a tangible user interface. This PoC incorporates a Peltier module to provide localized thermal stimuli (hot or cold) on the user's fingertip, a temperature sensor for skin temperature measurement, and a display. The core innovation lies in using this touch-based thermal sensation as an immediate feedback modality.

The methodology involved two user studies:

  1. Preliminary Study: This involved calibrating thermal models (Zhang et al.'s local thermal sensation model and Rabbani et al.'s adjusted Predicted Mean Vote model for global thermal sensation) and refining the experimental protocol.
  2. User Experiment: A within-subjects experiment was conducted where participants controlled the PoC thermostat in two conditions: (a) without touch-based thermal feedback (baseline) and (b) with touch-based thermal feedback. Participants were asked to find their ideal temperature setpoint. Data collected included quantitative measures (setpoint adjustments, time) and qualitative data (think-aloud protocols, questionnaires, interviews).

The thermal models were crucial for translating desired thermal comfort levels into tactile sensations and for predicting global thermal sensation based on local touch.

Datasets

Not discussed in the paper. The study relied on experimental data collected from participants interacting with the developed prototype.

Key Insights and Ideas

  • Touch as an Intuitive Thermostat Interface: The research strongly suggests that touch-based thermal sensation can serve as a natural and intuitive modality for controlling thermostats, bypassing the complexities of traditional interfaces.
  • Bridging the Feedback Gap: Immediate tactile thermal feedback effectively bridges the "delayed feedback" problem by providing users with a direct, real-time indication of their setpoint adjustments and anticipated comfort.
  • Anticipation and Strategy Shift: The presence of touch-based feedback significantly altered user strategies, shifting them from reactive adjustments to more anticipatory control, leading to fewer setpoint changes.
  • Local to Global Sensation Mapping: The study demonstrates that localized thermal stimuli on the fingertip can be effectively mapped to perceived global thermal sensation, providing a useful proxy for room comfort.
  • Improved User Experience: Participants found the touch-based system easier and more accurate for setpoint adjustment, leading to higher satisfaction and a better perceived match between their actions and the resulting comfort.

Key Results

  • Reduced Setpoint Adjustments: In the session with touch-based feedback, participants made significantly fewer setpoint changes (average of 1.38 vs. 3.31 in the baseline session).
  • Avoided Overshooting: Participants were more conservative with their initial setpoints and showed less overshooting behavior when using touch feedback.
  • Higher Satisfaction with Mapping: A significant majority of participants (9/13) reported very high satisfaction with the mapping between touch-based and global thermal sensation.
  • Increased Perceived Ease and Accuracy: Participants found the setpoint adjustment task "much easier and more accurate" with touch-based thermal sensation.
  • Reduced Visual Influence: The influence of visual feedback on setpoint selection significantly decreased when touch-based thermal feedback was available.

Limitations and Future Work

  • Real-World Complexity: The authors acknowledge that the complexity and multifactorial nature of thermal perception might pose challenges in real-world, dynamic environments.
  • Individual Differences: While calibration was included, further improvements to thermal models and calibration processes are needed to better accommodate individual differences in thermal perception.
  • Single-User Scenarios: The current study focused on single-user scenarios. Future work should explore multi-user scenarios where thermal comfort preferences may differ, and touch-based modalities could act as mediators.
  • Alternative Modalities: The researchers suggest exploring alternative locations (e.g., air) and modalities for delivering thermal stimuli.
  • Energy Consumption: While improved comfort and control are demonstrated, further investigation into the direct impact on energy consumption is warranted.

Summary updated

BibTeX
@inproceedings{huang2024making,
  author = {Yidong Huang and Gaëlle Calvary and Yann Laurillau},
  title = {{Making Thermostats Great Again: Investigating Touch-Based Thermal Sensation for Manual Control of Thermostats}},
  booktitle = {AVI},
  year = {2024},
  doi = {10.1145/3656650.3656672},
}
Figure