iCity 2: ICT4iCity project

Information and communication technology (ICT) is an important component of the smart city. Thematic areas addressed by ICT4iCity included data centers for urban data collection and storage, (mobile) networks, new information services for citizens, building and environmental monitoring with sensors, and GIS data for urban construction projects.

Urban data ranges from geospatial data for environment and construction to geometry data of buildings and information from the Internet of Things. How can a city's heterogeneous data sets be made available in a uniform manner via open interfaces and standards? Previous communication networks for data transmission are currently being expanded by new radio technologies such as 5G. What potential do these technologies have for use in buildings and cities? The increasing volume of data also requires automated processing to make the data valuable. How can new methods such as deep learning contribute to this? Finally yet importantly, how can data collected in the city and its buildings contribute to a better use of public resources, for example via new services for citizens?

In four subprojects, ICT4iCity and its partners from industry and the public sector were researching on various aspects and technological approaches to the before mentioned questions:

1. Urban Data Platform (Prof. Dr. Volker Coors)
2. Smart Public Buildings and Infrastructures (Prof. Dr. Dieter Uckelmann)
3. 5G-4-iCity (Prof. Dr. Dieter Uckelmann)
4. Deep Learning for 3D Geometry Data with Application in Classification of Building Types (Prof. Dr. Nicola Wolpert)

With the Urban Data Platform, a data infrastructure was developed, that allows uniform access to a heterogeneous dataset via open interfaces of the Open Geospatial Consortium, taking into account 2D and 3D geobase data as well as sensor data (1).

At the interface from the public building to the city, new IoT applications for municipalities and cities based on open software and open standards were investigated and tested using a basic installation (2)

The potential and technical basis for the use of 5G in buildings and for geolocation were determined; furthermore, a 5G evaluation sensor board was assembled (3). For the automated classification of building types in digital 3D models, existing neural network architectures (Deep Learning) were adapted and further developed (4).