Optimization of the Characterization of the Thermal Properties of the Building Envelope: Analysis of the Characterization of the Façades using Artificial Intelligence

How much energy a building consumes often comes down to what we cannot see: the thermal behavior of its envelope. In this SpringerBrief, David Bienvenido-Huertas and Carlos Rubio-Bellido tackle a persistent challenge in building physics—how to characterize the thermal properties of façades accurately and efficiently. Their answer lies in artificial intelligence.

Why Thermal Characterization Is a Bottleneck

Traditional assessment methods, whether theoretical or experimental, can be time-consuming, invasive, or limited in scope. Yet reliable thermal data is essential for energy retrofits, regulatory compliance, and the broader goal of reducing building sector emissions. The authors frame this problem within the context of low-carbon economy targets and the pressing need for nearly zero-energy buildings.

Artificial Intelligence Steps In

The core of the book introduces artificial intelligence algorithms—specifically artificial neural networks and multilayer perceptrons—as tools for estimating thermal properties from available data. The authors explain the methodological framework, including dataset generation, training, and testing processes, making the approach accessible to readers with a background in engineering or building physics.

What the Book Covers

• The influence of envelope thermal properties on building energy performance
• Regulatory frameworks and energy efficiency goals
• Static and periodic thermal property assessment methods
• AI algorithms and dataset preparation
• Estimation of stationary and periodic thermal properties
• Comparative analysis of AI-based estimation performance
• Error elimination in the thermometric method using multilayer perceptrons
• Determination of the constructive period of buildings

From Stationary to Periodic Properties

Separate chapters address the estimation of stationary and periodic thermal properties, highlighting how AI can handle both steady-state and dynamic thermal behavior. The comparative analyses provide insight into the strengths and limitations of the proposed techniques, helping readers understand when and how to apply them.

Practical Applications and Error Correction

The final chapter extends the discussion to other experimental approaches, showing how AI can improve the thermometric method by eliminating errors and even help determine the constructive period of a building. These applications demonstrate the practical value of the research for real-world building assessment.

Who This Book Is For

This concise volume is aimed at researchers, building engineers, energy assessors, and professionals in construction and building physics who want to understand how artificial intelligence can streamline and improve thermal characterization. It is also suitable for graduate students exploring the intersection of AI and sustainable building practices.

Available as a digital ebook from Digital Delights.