Artificial Intelligence based tool for decision-making in urban stormwater management

Abstract

Artificial Intelligence (AI) is rapidly evolving and demonstrating potential for climate change adaptation in urban planning, including urban stormwater management. Through the case study of InflowGo, an AI-based stormwater model, this research identifies and appraises potential opportunities and challenges that such a tool can provide for decision-making in urban stormwater management. By the means of semi-structured interviews with urban water professionals, it was revealed that conventional decision-making, i.e. without AI application, is complex due to technical and organisational challenges, as well as insufficient speed, flexibility, and collaboration. AI has the potential to address these challenges and fill these gaps as it can accelerate decision-making, both technically and by facilitating the selection of different alternatives earlier in the process. It can also foster interdisciplinary collaboration and participation of stakeholders in decision-making by running the models during meetings, helping to break silos, and supporting educational purposes. To do this, the user-friendliness and web-based nature of the tool were identified as two contributing success factors. Although appearing as potentially relevant, AI was found not capable of overcoming the complexity and uncertainty of the field, the world, and climate change. Moreover, it can bring new challenges and shortcomings related to the ethical responsibility of making decisions, resistance to change and fear of the unknown, as well as legal and cybersecurity aspects. Further research is encouraged to investigate the application of AI-based tools in different contexts and with fully developed products, and to explore the potential educational added value of AI in decision-making.

Popular Abstract

Artificial Intelligence has the potential to revolutionise urban stormwater management decision-making to make cities ready for challenges like climate change, urbanisation, and ageing infrastructure. Using the case study of the AI-based stormwater model InflowGo, the study describes the shortcomings of the current decision-making process, and suggests potential benefits of using AI to fill these gaps. Furthermore, it discusses AI’s potential limitations and arising challenges that come with its application. The research is based on the case study of the InflowGo tool, still in development, and 16 semi-structured interviews conducted with urban water management professionals from Germany, Denmark, and Sweden.

Decision-making in urban stormwater management is a complex process that involves politics, teamwork, and technical know-how. It requires experts to use conventional tools for modelling and simulation, making it a time-consuming and repetitive process. Moreover, this process involves extensive data analysis and coordination among many professionals from different departments, which is found to be lacking flexibility and in fact non-inclusive. The research highlights several potential benefits of integrating AI into stormwater management. AI's ability to rapidly analyse large datasets can significantly reduce the time required to develop and implement effective stormwater strategies. By simulating various stormwater scenarios in real-time during meetings, an AI-based tool can allow urban stormwater managers to visualise potential impacts, compare different solutions, and make informed decisions more quickly and efficiently than before. This not only speeds up the decision-making process but also encourages participation and collaboration among stakeholders from different fields. Furthermore, InflowGo’s user-friendliness and web-based design make it accessible to various stakeholders without the need for specialised training or powerful computing resources, and at a lower cost. This accessibility coupled with the quicker nature of AI fosters even greater participation and interdisciplinary collaboration, making the decision-making process more inclusive while also offering an educational added-value to the tool’s application. Despite these advantages, the study also acknowledges the limitations and new challenges introduced by AI. While AI-based tools could streamline many aspects of stormwater management, they cannot entirely eliminate the inherent complexities and uncertainties associated with climate change and urban stormwater management. Moreover, the introduction of AI raises various issues about ethics and the responsibility of decisions made, resistance to adopt new methods, lack of technological understanding behind AI, as well as legal and cybersecurity considerations. Finally, further research is encouraged to explore how similar AI-based tools perform in different settings and with fully implemented AI. Additionally, it is considered worth investigating the potential educational added-value that AI tools can have for training and learning purposes.

• author: Bilmes, Sofia and Berment, Antoine Baptiste Marie
• supervisor: Mo Hamza
• organization: Division of Risk Management and Societal Safety
• alternative title: Navigating tomorrow's rains: exploring the potential of AI in urban stormwater management
• course: VBRM15 20241
• year: 2024
• type: H2 - Master's Degree (Two Years)
• subject: Technology and Engineering; Social Sciences; Earth and Environmental Sciences
• keywords: decision-making, decision support system, modelling tool, Artificial Intelligence, Machine Learning, Deep Learning, disruptive technology, urban stormwater management, sustainable urban planning
• language: English
• id: 9158077
• date added to LUP: 2024-06-10 11:20:05
• date last changed: 2024-06-10 11:20:05