P8: Smart Building Technologies

Principal Investigator

Assoc. Prof. İpek Gürsel Dino

Institution(s): METU (Leader), JeoIT
Strategic Target Name: At different scales of the built environment (room, floor, building, group of buildings, neighborhood, etc.), developing prototypes of user-oriented digital platforms (artificial intelligence-supported control algorithms and decision-support systems) that reduce building energy consumption and CO2 emissions.
Scope of Activity of the Project: Development of artificial intelligence algorithms, deep learning and computational intelligence algorithms; optimization, autonomy software, data analytics software, decision support systems, digital twins, game and interaction technologies, augmented reality, virtual reality.

Project Summary

Although the importance of increasing the energy efficiency of buildings, reducing CO2 emissions, ensuring user comfort in indoor environments and meeting energy needs with clean energy sources is known, there is currently no clear technology to express numerically the actions that will increase building and system performance, to determine the right decisions with computational methods and to finalize these decisions by sharing it correctly with users.

The main goal of this project is to develop digital platforms that reduce building-related energy consumption and CO2 emissions at different spatial scales and test them in different pilot regions. For this, visualization / interaction tools such as artificial intelligence- based analysis and control algorithms, IoT- based performance monitoring, physics-based building energy simulations, virtual / mixed / augmented reality will be used. All technologies to be developed in the project have already been developed at TRL 3 level at METU and JeoIT; In this project, it will be increased to TRL 6 – TRL 7 level.

The goals of the project are as follows:

Within the scope of the room-scale "Learning Smart Air Conditioner" application, monitoring of building users, assets and systems that consume energy (such as air conditioning, HVAC, electric vehicle charging, etc.) and provide energy (such as photovoltaics, etc.) in the building, by learning from the data obtained as a result of monitoring. Development of control algorithms that are energy efficient and enable personalization.
Within the scope of the "Urban-Scale Building Energy Analytics Platform", it develops artificial intelligence -supported decision-support and high-tech visualization tools for energy-efficient and climate-neutral building renovation processes. Climate change mitigation / adaptation actions at the city scale (with the support of the Ministry of Environment, Urbanization and Climate Change) and 'hourly' forecasting of the short-term impact (day-ahead forecasting) of building electrical energy on the electricity grid are aimed.
"Smart twin" application aims to provide a common data infrastructure to the first two applications. In the 'Digital Twin' study, it is aimed to provide a) data definition, b) sharing, c) analysis and d) presentation capacity of the room, building and urban scale with a common infrastructure.

Side-by-side comparison of a real laboratory shelf and its digital twin model

Top-down view of METU buildings in the digital twin with a room selection menu

Digital twin laboratory view with a data analytics panel

Project Team

Prof. Dr. İlkay Ulusoy

Researcher · METU

Research Areas: Signal processing, machine learning, computer vision and pattern recognition, probabilistic graphical models

Assoc. Prof. Sinan Kalkan

Researcher · METU

Research Areas: Artificial intelligence, computer science, computer vision and pattern recognition, engineering and technology

Assoc. Prof. Elif Sürer

Researcher · METU

Research Areas: Computer science, augmented reality, virtual reality, 3D displays, human computer interaction, engineering and technology, visualization and interactivity