DataScience4SmartQuarters - Energy saving potentials through neighbourhood and community planning – Part 2

In combination with high-resolution simulation data from the predecessor project SmartQ+ and data from large-scale studies, transferable test data and models for different types of space are created using various data science models (e.g. ML, regression analyses, etc.). This supports municipalities in strategic planning and in quickly and efficiently estimating the effects of their decisions.

Background and Motivation

Simulations in the areas of energy grids, buildings and transport are often time-consuming and resource-intensive, which makes evidence-based decision-making in municipalities more difficult. Changes in individual parameters require lengthy recalculations and their impact only becomes apperent after the simulation is complete. This makes it practically impossible to test all parameter combinations. A solution that can immediately estimate the effects of changes with sufficient accuracy would provide administrations with a better understanding and enable targeted formulation of optimised scenarios.

Goals and Innovation

DataScience4SmartQuarters is developing a proof-of-concept study (TRL3-4) for a user-friendly, fast, and cost-effective planning tool for municipalities. This tool will enable the rapid estimation of simulation results in the areas of buildings/energy (e.g., electricity consumption) and mobility (e.g., modal split and traffic generation) as a result of new developments or land-use changes.

The tool is designed to estimate the impact of parameters (renovation rate, heating replacement, PV expansion, e-mobility) in "real-time." This allows municipalities to make decisions quickly and cost-efficiently, helping them to achieve climate goals more rapidly.

In the developed method, when simulation parameters are changed, an initial estimation of the results is generated using Data Science methods (e.g., machine learning, regression methods). This allows interesting scenarios to be tested in advance before detailed simulations are conducted. New simulation data is incorporated into the improvement of local training datasets. Although this approach is discussed in the literature, its implementation in the municipal context has not yet been tested.

Methodological Approach

To ensure the national transferability of the results, large-scale data and studies, such as "Österreich unterwegs", will be analyzed with respect to spatial structures. Models will be developed and tested that can be applied to typical settlement structures and demographic patterns. Various models will be examined to link meaningful input parameters with relevant output variables.

The integration of existing databases and systems will be evaluated. The simulation data from SmartQ+ will serve as a valuable foundation and training dataset. The interdisciplinary collaboration of experts in spatial planning, data visualization, building physics, energy economics, and transportation planning enables a scientific comparison and testing of these methods. Workshops will involve stakeholders and future users in the development process.

Expected results

The project is aimed at decision-makers, planners, developers, and other stakeholders in neighborhood development. It offers a fast, cost-effective way to evaluate planning decisions and optimize scenarios before they are simulated in detail. The workflow can be used in participatory planning processes, as simulation results are available "at the push of a button."

Municipalities can independently run simulations and adapt them to changing conditions, enabling the development of targeted energy policies and faster decision-making. This promotes actions to reduce CO2 emissions. The project aims for widespread dissemination and application of the results.

Project management

Stefan Bindreiter
TU Wien, Institute for Spatial planning, Research Unit of Local Planning, Spatial Simulation Lab

Project or cooperation partners

• TU Wien, Institute of Material Technology, Building Physics, and Building Ecology, Research Unit of Building Physics
• yverkehrsplanung GmbH
• Energiepark Bruck/Leitha

TU Wien
Institute of Spatial Planning, E280-4
Stefan Bindreiter
Karlsplatz 13
A-1040 Vienna

Tel.: +43 (1) 58801 - 280430
Email: stefan.bindreiter@tuwien.ac.at
Web: www.tuwien.at/ar/simlab