Low-threshold modeling on a platform
Digital simulation models form the basis for many “what if” scenarios in urban digital twins. They transform data sets into insights, map correlations, and enable a wide range of possible futures to be generated in a short period of time, allowing more robust decisions to be made based on this information. However, creating simulation models is often time-consuming, requires specialized and expensive software, and necessitates extensive prior knowledge.
In order to create simple, transparent, and reusable simulation models with a direct connection to urban data for urban digital twins, the City Science Lab at HafenCity University has developed the Urban Model Builder as a freely accessible platform. With the Model Builder, both simple and more complex systems can be mapped, simulated, and shared with other users—directly in the browser. Thanks to the graphical clarity of the user interface, intuitive handling, and minimal input of values and parameters, a wide variety of sectoral models can be collaboratively created and used in a short amount of time. These include population forecasts, mobility models, resource flows, and many other urban systems.
Open standards, real-time collaboration, and application examples
The Urban Model Builder is based on the open-source simulation library simulation. It offers the possibility to create both system dynamic models and agent-based models, as well as to try out, test, and further develop models published by other users. The various building blocks available to users can either be filled with their own assumptions and values or linked to real data sets, e.g., from the Urban Data Platform Hamburg. By simply placing the graphical components on the canvas and connecting them, logical relationships and calculable schemata are created, which can be versioned and persistently archived on the one hand, and published for other users of the tool on the other. In order to design the models collaboratively, it is also possible to work on the model with other users in real time.
The following population model of Hamburg can be cited as a simple example of its application. Consisting of only four components, it is possible to create a simulation model for experimentation. The “current population” as a stock receives its data set on the population density of the individual districts via the standardized OGC API Features interface from the Urban Data Platform Hamburg. Growth as a flow depends on a specific growth rate and the population at a specific point in time. As a global setting, time intervals can be defined at which the “current population” of the individual districts can be simulated. The predicted population data at specific points in time can now be read in a graph.
In Practice: A Population Forecast in the Urban Model Builder
Provision via OGC interfaces and open-source development
The models created can also be made available for other applications via an OGC API Processes interface on the Urban Model Platform (link) after review and approval.
The Urban Model Builder is an open-source development that was implemented in collaboration with the company triggercode. It is completely free to use.