Rebooting the underpass, bridging districts
A co-designed, AI-visualized underpass prototype shaped by citizen insights to enhance safety, accessibility, and social inclusion.
Challenge
How can citizens be involved in a design thinking process to make a pedestrian underpass barrier-free and safe?
The challenge revolves around the redesign of the Luisentunnel. However, the core problem is to create a public participation process that is truly inclusive. Municipalities face the issue that often the same social group, and foremost the exact same people, always attend most of the public participation processes.
Team
Shahzaib Hameed, Bui Nguyen Ngoc Huyen, Tianzhuo Wang, Andreas Ruthus
Contact: shahzaib.hameed@tum.de
About the prototype
We approached the Luisentunnel redesign challenge through a three-phase process. In Phase 1, we collected citizen feedback using digital and printed surveys created with JotForm AI. These multilingual, user-friendly forms were shared via QR codes, WhatsApp, and in public spaces to reach diverse groups—parents, students, immigrants, elderly, and people with disabilities. In Phase 2, we compiled the responses and used ChatGPT and AI tools to analyze the data, identifying common themes like safety, lighting, accessibility, greenery, and cleanliness. These insights were transformed into text prompts describing an ideal underpass, which were then input into AI design tools like Midjourney to generate visual concepts. Phase 3 focused on summarizing feedback with tools like word clouds and mood boards. We produced 3D AI-generated visuals that directly reflected citizen input, bridging the gap between public feedback and design outcomes. This process demonstrated how AI can enhance participatory urban planning in a tangible, inclusive way.
Outputs
Project Report of Team 2 Fürth
Impressions
