Beyond - Virtual Reality enabled energy services for smart energy systems

Collaborative R&D project to develop the next generation energy services with the interplay of various technologies: Virtual Reality (VR), machine learning, physical simulation and Internet of Things (IoT) platforms.

Short Description

Starting point / motivation

The Austrian government is committed to accelerating the transition of the energy system and achieving CO2 neutrality by 2040. To achieve this goal, Austria must significantly intensify its efforts to decarbonise all parts of its energy sector. Buildings account for about a third of the total end-use energy demand.

The government plans to phase out oil and coal heating systems by 2035 and limit the use of natural gas for heating in new buildings from 2025. Europe is now entering the fourth wave of energy efficiency as characterised by the increased digitization of society, distributed energy resource (DER) deployment, and the changing nature of the energy supply and demand mix.

Intelligent energy services such as predictive maintenance, demand-side management and model predictive control are central components for reducing the energy consumption of buildings and transforming buildings into intelligent actors in the next generation of smart energy systems.

Demand-side adoption of IoT technologies in homes, commercial and industrial buildings, as well as educational and community facilities offers enormous potential to increase efficiency through more comprehensive energy management systems.

Contents and goals

BEYOND aims to develop the technological foundation for "Next Generation Energy Services", which is made possible by the interplay of the following technologies:

  • Virtual Reality (VR) for visualization and real-time interaction with the real building;
  • Machine learning and physical simulation to show the real-life effects of interventions and decisions;
  • Internet of Things (IoT) platforms for the realisation of smart systems and bidirectional real-time communication between the building and its users.

The technological developments will be tested and evaluated on the basis of two use cases "Human Aspects in Buildings" and "Predictive Maintenance and Error Diagnosis".

Methods

The first use case "Human Aspects in Buildings" will allow understanding of the impact of user decisions on the performance of buildings. For this purpose, white-box models will be derived from existing BIM models in combination with black-box models forecasting the energy consumption in buildings and building simulations to evaluate the daylight levels (including different levels of glazing and shades).

The models will rely on a number of monitored variables: 

  • energy use
  • CO2
  • humidity
  • air temperature
  • climate data
  • daylight
  • PV electricity
  • etc. 

These models will be integrated with VR technologies to strengthen users' sense through visual representations (e.g. of heat loss, heat flux, daylight etc.), acoustic (e.g. acoustic warnings when temperatures or low daylight levels are exceeded etc.) and haptic (e.g. to feel temperature changes etc.) information as well as providing real-time evaluation of the users' decisions.

The second use case "Predictive Maintenance and Error Diagnosis" will use monitored data to develop machine learning models for predictive maintenance aiming at predicting at an early stage when a system (e.g. HVAC systems etc.), machine or any of its components is likely to fail in order to perform timely maintenance before a fault occurs. For this purpose, the models will use considerable amounts of monitored data indicating potential signs of a system (or component) failure and their reliability: 

  • physical health aspects of the equipment,
  • machinery or components (e.g. pressure, vibration, temperature, viscosity, acoustics, flow rate, etc.),
  • energy data (e.g. electricity and gas consumption) and
  • environmental data (e.g. indoor temperature and relative humidity). 

The use case will also adopt machine learning methods, integrity factors (e.g. visual aspects, wear, etc.), statistical reference approaches and other statistical techniques based on the historical data for automated error diagnosis. These models will be also integrated with VR to aid on-site maintenance diagnostic procedures by providing enhanced visualisation, supporting and resolving early-stage problems (i.e. before failure) and identifying acute errors.

Expected results

The development of the two use cases in the context of VR will lay the foundations for research into interactive machine learning methods and user-friendly building optimisation mechanisms and maintenance of systems. This will provide the know-how for building planning, but also for the use of existing buildings.

Findings of the BEYOND project will benefit, amongst others, innovative companies in energy and fault detection services, building management and automation, simulation software and VR technology.

Policy decision-makers and end-users will also benefit from enhanced ways of interacting with these "Next Generation Energy Services" that integrate real-time interaction via VR, online simulations, and access to real-time data and historical data via the IoT platform.

Project Partners

Project management

Institute of Building Physics, Services and Construction (IBPSC), Graz University of Technology

Project or cooperation partners

  • Institute of Interactive Systems and Data Science (ISDS), Graz University of Technology
  • Institute of Software Technology (IST), Graz University of Technology
  • EAM Systems
  • EnaLytics

Contact Address

TU-Graz
Institute for Building Physics, Building Technology and Building Construction
Univ.-Prof.Dipl.-Ing.Dr. Christina J. Hopfe
Lessingstraße 25/III
A-8010 Graz
Tel.: +43 316 873 - 6240
E-mail: c.j.hopfe@tugraz.at
Web: www.tugraz.at/en/institutes/ibpsc/