Non-invasive assessments of thermal discomfort in real time

• People make distinctive gestures or movements when they are thermally uncomfortable, for example self-hugging when uncomfortably cold or brow-wiping when hot. Extreme thermal conditions reinforce this tendency. These gestures may be affected by various competing motivations such as emotional or physiological responses and cultural traditions. Several software applications can now identify and track movements of a person’s skeletal joints or keypoints in real time; these include hands, arms, elbows, head, etc.. A procedure was created to identify gestures related to thermal discomfort and then to decide if a person is uncomfortably warm or cold. When a discomfort-related gesture is detected, it is scored based on the type of gesture and recognition confidence. This score is fed into a “Thermal Comfort Index” (TCI). A zero TCI corresponds to thermal neutrality and higher positive or negative values correspond to feelings of warmth or cold, respectively. The TCI diverges further from zero when the detected gestures are frequent or intense. A key feature is a “library of gestures”; the procedure consults this library to determine if a particular gesture might be associated with thermal discomfort. This method was applied to a single person but could also be applied to large groups. This method of tracking thermal discomfort is well suited for locations where occupants are not easily able to communicate thermal preferences.

Subject headings

TEKNIK OCH TEKNOLOGIER  -- Samhällsbyggnadsteknik -- Husbyggnad (hsv//swe)

ENGINEERING AND TECHNOLOGY  -- Civil Engineering -- Building Technologies (hsv//eng)

Keyword

thermal discomfort

non-invasive measurement

gesture

Kinect

HVAC

Publication and Content Type

ref (subject category)

kon (subject category)