ISCN Online Advisory Programme #15: Green infrastructure in Smart Cities

Summer 2024 was the world's hottest on record globally. Climate change is unfolding and the big and urgent challenge is to limit it and at the same time begin with adaptations. At the same time, trees and green infrastructure can cool the surface temperature in cities by up to 12%! Next to this cooling effect, green infrastructure has many other striking beneftis for the environment such as habitat for biodiversity, protection for erosion, improvement of air quality and many more. As such, green infrastructure is one of the backbones of urban resilience and quality of life. In striving for common good-oriented smart cities, it thus has to be one of the absolute focusses - even more so as green infrastructure and tree populations themselves are under heavy distress from diseases and climate change. 

This Online Advisory Programme therefore made a deepdive into how green infrastructure in smart cities can be supported by technology. Great speakers presented respective measures, with "Arból IoT", a solution in Mexico, following "BAKIM", a solution from Bamberg in Germany, a Smart City Model Project (MPSC). Sascha Götz from the city of Bamberg and Jonas-Dario Troles, technical project lead of BAKIM from the Otto-Friedrich-University explained how the municipal administration and the university in Bamberg developed together and autonomously an AI-based solution for monitoring the health status of trees in Bamberg's surroundings with imagery from drones. Sascha highlighted the success of this constellation: 

This is not a piece of work of industry, this is a piece of work of university and public administration, in a very practical manner, on a very high level of scientific intelligence 

He also contextualised the project in the municipal administration, in explaining how the divisions of arborists and foresters have different goals and needs when it comes to tree management and how the project BAKIM can cater to both. 

Jonas guided the audience through some technical intricacies of training AI models that can process and interpret imagery from forests. Measurement can be subdivided in the dimensions of counting the trees, tracking their vitality and identifying their types. Especially the latter is difficult for algorithms if unsupervised and requires a lot of prior human labeling of trees as "ground-truth data". With BAKIM, Bamberg managed to publish the worldwide largest dataset of such high resolution, tree amounts and quality of labelling, in addition to a high degree of accuracy in classification, or identification of pests, such as mistletoes. (The respective paper can be accessed via https://doi.org/10.3390/rs16111935)

 

Now the project is developing further contacts with other municipalities interested in cooperation in order to gain access to a further extended set of labelled tree data and in return scale the usage of their solution. Interested parties can also step into contact with the project via the International Smart Cities Network at iscn@giz.de.
 

Árbol IoT

Arturo Cadena Hernandez from GIZ then gave a keynote on "IoTrees" (“ Árbol IoT” in Spanish), a project from Guadalajara, Mexico, implemented by GIZ on behalf of the Federal Ministry for Economic Affairs and Climate Action within the framework of the German international climate initiative (IKI) and in cooperation with the Federal Ministry for Housing, Urban Development and Building. The project built on the objective necessity of dealing with increased temperatures and extreme weather events but also specifically with the citizenry expressing high priority for urban trees through a participatory budgeting process. As a cross-platform it aimed at both establishing more information on the status quo of the trees in Guadalajara through an urban tree inventory and at translating this crucial baseline into new planning and inclusive management practices of trees. 
The inventory and mapping of trees was done in a complementary hybrid form, employing technical methods, such as LiDAR for exact parameters (Green canopy coverage, tree location, quantity of objects, built environment, etc.) and participatory field work for information on diseases, tree species, photografic validation, etc. Participatory fieldwork was enabled in combining a mobile app, low-cost DIY environmental sensors for selected trees and a web application. The ca. one million trees of Guadalajara were spatially mapped and additional data on them collected through crowdsourcing from citizens by using the prepared app. At the same time, the app provided additional information on the trees  to further raise ecological awareness and gamified the valuable recording work of citizens. 
 

The implementation of Árbol IoT is documented in detail for replication: 
 

For further questions, feel free to reach out to iscn@giz.de