Jobs

If you are interested in doing your thesis with us or work on any of the topics as a trainee, please use the contact form to get in touch. Language is English or German. 

Bachelor Thesis: green roof classification

In order to implement suitable adaptation strategies, it is necessary to know some information about the roofs of the buildings in question. 

On flat roofs and roofs with not much slope, green roofs can be installed and function as a buffer for rain water during heavy rain events.

In your thesis, you will: 

• review all publicly available datasets on roof shapes for Germany
• review the state of the art on green roof classification and pick a promising one
• implement a script that takes the address of a property as a input, uses the chosen green roof classification library and exports:
• the total roof area
• the roof area listed as consecutive areas with the same slope and the direction the slope faces (north, southwest, …) 
• the percentage of roof that is already a green roof
• the area of roof that could be turned into a green roof (neglecting the structural constraints)
• calculate how much rainwater could be buffered under different implementation scenarios (e.g. extensive greening on flat roofs only, all roofs greenes, etc.)

Bachelor Thesis: Counting single trees

Trees influence the micro climate of a neighborhood. They break the wind, shade the area from the sun, evaporate and hold back water. Currently, there is no comprehensive dataset on where trees are growing within a city. However, there are quite a few AI tools to classify trees from aerial and satellite images, photogrammetry based 3D data or laser scans.  

In your thesis, you will: 

• Review data sets for urban areas that have information on trees
• Review the state of the art of tree classification in urban areas. 
• choose the 3 most promising classifiers 
• use them to classify 3 test areas
• evaluate their performance, which includes gathering ground truth data
• evaluate their sustainability based on their energy consumption

Master Theis: Unreal Rain Water Management

Hypothese: Mit der Unreal Engine (ue5) lässt sich hinreichend genau simulieren, welche Auswirkungen die Umgestaltung eines Außengeländes auf den Abfluss von Regenwasser bei Starkregen hat.

Die Story dahinter: wir unterstützen Nachbarschaften, soziale Einrichtungen und andere Gebäudeeigentümerinnen dabei, Klimafolgenanpassungsstratiegien zu entwickeln. Eine Folge des Klimawandels ist die Veränderung der Niederschlagsmengen. Starkregen wird der neue „normale“ Schauer. Mit einer guten Anpassungsstrategie ist dieser Wandel handhabbar. Und um zu testen, ob eine Anpassungsstrategie gut ist, wollen wir simulieren, was bei einem Starkregenevent passieren wird. 

Das machst du in deiner MA:

• baue eine Pipeline, um real world Daten der Geländebeschaffenheit, Bodeneigenschaften, Bebauung, Materialien und Vegetation in UE zu importieren.
• Simuliere in UE Starkregenevents an einem ausgewählten Standort. Berücksichtige dabei, wie sich Wasser auf verschiedenen Materialien verhält.
• implementiere eine Möglichkeit, alle Randparameter und Ergebnisse der Simulation, wie z.B. die Wasserstände nach diesem Starkregenevent aus UE zu exportieren. 
• validiere die Simulationsergebnisse durch eine Messreihe an dem realen Standort

Master Thesis: Photorealistic Object Insertion

At Reimagine Spaces, we do believe that adaptation measures work the best when they are co-creatively designed by a group of people that includes those who will use the reimagined space and experts from various fields, such as landscapers, architects or fire fighters. 

When asked to come up with ideas to change the current state of a property, be it the outdoor area of a school or the garden of an elderly home, some groups are stricken by “blank page fear” and some individuals who are not in the habit of being creative have a hard time getting involved.  The creative process can then be greatly enhanced and facilitated by providing good examples, shown visually as pictures. The most impact is achieved if the image shows what the actual area would look like after adaptation measures are implemented, for example if the concrete surface was removed and replaced by a group of trees, the plain grass turned into a pond for rainwater management or shaded seating areas put up. 

With current advances in AI powered Photorealistic Object Insertion, this process can be automated and no longer needs hours of work done by a 3D artist / landscape architect. Through natural language prompting, even inexperienced users are able to try out different adaptation variants. 

In your thesis you will:

• compare the currently available object insertion AIs
• design and test a workflow on how to use this tool in workshops with diverse groups of people. This might include drafting suitable prompts or fine tuning an AI model 
• provide a guide for other facilitators on how to use it in their workshops