Skip to main content Skip to secondary navigation

A Socio-technical Approach to Increasing the Use of Natural Cooling in Residential Buildings

Main content start
title_picture_a_socio-technical_approach.jpg

Research Team

Our Motivation:

“Natural methods of cooling residential buildings are increasingly critical in urban areas. This is particularly true  in mega-cities in emerging markets where high temperatures and wide income distributions make efficient and resilient mechanical cooling inaccessible. However, designing naturally cooled buildings relies on an understanding of occupant behavior and the relationship between the indoor and outdoor climate. The need for robust data thus makes designing accurate models difficult.”


 

Icon Contribution

Research Contribution

Our key contribution is an approach for predicting occupant behavior using inexpensive datasets. This approach can be used to tailor the operation of existing buildings and design of new buildings for improved performance while ensuring more reliable results. 

We also present novel methods of assessing the accuracy of methods for predicting occupant behavior. 

Icon Problem

Problem

Practical Problem

Existing models for predicting window opening and closing rely on long periods (1 year or more) of labeled data which are incompatible with the rapid iterations in building development and management and are often inaccessible to a typical practitioner. 

If we hope to move toward a world where sustainable buildings are the norm, and not an exception for expensive, high-visibility projects, developing cheaper methods for predicting and understanding occupant behavior is essential. 

Conceptual Problem

Current methods of predicting occupant behavior rely on data that is inaccessible to practitioners.

Icon Solution

Solution

Develop a practical method for predicting occupant window operation using low amounts of data, namely a comparison between indoor and outdoor temperatures.

Icon added value

Added Value For The Industry

Practitioners are able to bridge the performance gap between the predicted and actual performance of buildings through the ability to rapidly develop a better understanding of occupant behavior.

Icon Cooperation partner

Cooperation Partner

---

Icon Timeline

 Timeline

DateActivityOutcome

Year 2022

Research became awarded: A Socio-technical Approach to Increasing the Use of Natural Cooling in Residential Buildings

 

Aug 2022

Obtain data from building on campus

 

March 2023

Develop model for predicting window opening on campus

 

July 2023

Submit paper to conference

 

 

 

Project Summary

(Provides you with a brief and clear summary of the insights and outcomes at the end of the funded year.)

Natural methods of cooling residential buildings are increasingly critical in urban areas. 

However, designing naturally cooled buildings relies on an understanding of occupant behavior as well as the relationship between the indoor and outdoor climate. The need for robust data, which is often inaccessible during tight design cycles, makes creating accurate models difficult. 

Our work focused on developing a practical, unsupervised classification method for predicting occupant window operation (which is critical for naturally cooled buildings) using low amounts of data. 

We found that our developed method had comparable results to a general purpose machine learning model that has been widely used for classification problems, demonstrating its effectiveness for our described setting.

Contact Person

If you want to participate in the project please reach out to Juliet Nwagwu Ume-Ezeoke.