Interview

This is the second interview in our series on the innovative De-Cist project. The Developing Energy Communities with Intelligent and Sustainable Technologies (DE-CIST) is led by Dr. Rebecca Moody (EUR) and unites Erasmus University Rotterdam, TU Delft, Institute of Housing and Urban Development Studies, Resilient Delta Initiative, Erasmus Centre for Data Analytics, and the City of Rotterdam. Supported by ICLEI Europe through a 1 million Euro Google.org grant, we speak with Pablo Morato, a postdoctoral researcher at TU Delft, about an AI-powered tool helping Rotterdam reduce building energy consumption.

The De-Cist project and Morato’s role

The De-Cist project aims to improve energy efficiency in urban areas, specifically focusing on retrofitting Rotterdam’s buildings to reduce energy consumption. The project collects data on individual buildings and combines it with meteorological, air quality, emission, and socio-economic data to create comprehensive profiles. “Many older buildings have poor insulation, leading to excessive energy use for heating and cooling,” explains Morato. “Our goal is to develop an AI tool that helps decision-makers—municipalities, building owners, or citizens—identify the best retrofit solutions.”

As a postdoctoral researcher at TU Delft, Morato’s primary responsibility is developing this AI decision support tool. Working in a team—Seyran Khademi, Anna Maria Koniari, and Charalampos Andriotis— they have created a system that analyzes building data—including typology, age, and materials—to suggest appropriate retrofit solutions like enhanced insulation or upgraded glazing. The ultimate aim is to make energy-efficient renovations more accessible and easier to implement across Rotterdam’s diverse building stock.

From naval engineering to urban energy efficiency

Morato’s academic journey began in Spain with naval engineering, specializing in structures and hydrodynamics. His career then shifted to offshore wind energy before focusing on energy consumption. “Even though this is a new field for me, many computational methods I used before, such as optimization algorithms, are applicable to this project as well,” he notes.

How the AI tool functions

The tool integrates multiple data sources, including existing building records and energy efficiency information. It evaluates a building’s current energy demand and predicts changes after implementing different retrofit solutions. The AI then optimizes choices by balancing energy savings with economic and social factors.
“We focus mainly on the building envelope—the exterior walls, roof, windows, and floors,” Morato explains. “For example, we analyze different types of insulation or glazing options for windows.”

Challenges in development

Data access and management proved to be significant hurdles. “The Netherlands has a lot of open data, which is great, but it’s spread across different sources,” Morato says. Some sensitive data, like aerial building images, required physical transfers at municipal offices. “Finding and structuring the right data was actually a bigger challenge than the AI modeling itself.”
Another challenge was incorporating socio-economic factors into the decision-making process. “Energy efficiency isn’t just a technical problem—it’s also a financial and social one. But data on socio-economic aspects isn’t always available at the building level, so we had to find creative ways to incorporate it.”

The transdisciplinary experience

Working with social scientists was new for Morato. “At first, communication was a challenge—we speak different ‘languages’ in framing problems,” he reflects. The collaboration changed his approach: “In engineering, we rely on numbers and optimization models. But social scientists showed us that real-world decisions aren’t purely numerical.” This insight led to practical improvements like adding budget constraint features.

Adapting to user needs

The project required flexibility. “In most technical projects, you define requirements upfront. Here, we are constantly adapting,” notes Morato. Based on survey feedback, the team enhanced user-friendliness and financial realism. “That iterative process isn’t common in technical research, but it made our tool more practical.”

Future aspirations

Morato hopes the tool becomes a valuable resource for municipalities and citizens seeking to improve building energy efficiency. “More broadly, I hope this project serves as a model for transdisciplinary research. Working across fields isn’t always easy, but it produces better, more impactful solutions.”

The experience has influenced his perspective. “This project has influenced the way I present my research approach,” he concludes. The DE-CIST project exemplifies how data science and artificial intelligence can be applied with social context to address pressing urban sustainability challenges in ways that are both technically sound and socially relevant.