SELF²B - self-aware, self-diagnosing buildings, HVAC, and PV systems for the next generation of energy efficient operations

SELF²B develops and demonstrates an AI-based, self-learning, and self-diagnosing fault detection and diagnosis (FDD) solution for HVAC and PV systems in two buildings in Vienna. The innovation surpasses the current state of the art by combining semantic data, ontologies, and machine learning. The goal is to achieve energy savings and efficiency improvements in building operations and to make the technology widely applicable.

Short Description

Starting point/Motivation

To achieve the national and European climate neutrality goals by 2050, CO2 emissions must be significantly reduced. The building sector plays a central role in this: in the EU, it accounts for 40% of final energy consumption and 36% of emissions.

However, continuous, systematic monitoring of building operations is rarely carried out due to the complexity of heating, ventilation, and air conditioning (HVAC) systems, insufficient data and tools, or a lack of personnel, even though ongoing monitoring of operational parameters has great potential. Studies show that optimized operations and intelligent monitoring in the non-residential building sector could yield energy savings of up to 30%.

However, commercially available software for operational optimization requires highly qualified specialists, is complex to set up, or often necessitates an upgrade of the existing HVAC measurement technology in advance. The possibilities and advantages of intelligent fault detection and automatic fault correction in buildings have yet to be realized in practice.

Content and goals

SELF²B demonstrates an AI-based, self-learning, and self-diagnosing fault detection and diagnosis (FDD) software to optimize complex non-residential buildings with building automation systems. In addition to the heating, ventilation, and air conditioning (HVAC) systems considered in the project, photovoltaic systems are also included.

Methods

The innovations planned in the SELF²B project go beyond the current state of the art: The combination of semantic data and ontologies, heuristics, and machine learning guarantees scalable and robust solutions for HVAC and PV systems.

The project's planned combination of semi-supervised machine learning models with autoencoders, combined with automated clustering and classification models, also represents an innovation in machine learning that can potentially be applied to other areas.

The planned integration of users in the development process, as well as the focus on explainability and user-friendliness, addresses the relevant market hurdle of technological scepticism among key stakeholder groups for fully automated software solutions. Important research comes mainly from China and the USA, making this pilot project one of the first real-time implementations of its kind in Europe.

Expected Results

The solutions developed in the project will be demonstrated in the form of a software prototype in real-world operations within pilot buildings of the Federal Real Estate Agency and evaluated using a matrix of technical, economic, and ecological criteria.

Additionally, a technology concept for "self-learning, self-optimizing" existing buildings will be developed for the next generation of efficient building operations.

Project Partners

Projektleitung

TU Wien, Institut für Hoch- und Industriebau

Projekt- bzw. Kooperationspartner:innen

DiLT Analytics GmbH
Bundesimmobiliengesellschaft (assoziierter Partner)

Contact Address

TU Wien
Institut für Hoch- und Industriebau
Karlsplatz 13/210
A-1040 Vienna
Tel.: +43 (1) 588 01 215 21
E-mail: gerald.schweiger@tuwien.ac.at
Web: www.tuwien.at/cee/hib