A Good System for Smart Cities

A key goal of smart cities is to use multiple, large-scale datasets to make better, long-term urban development decisions for residents, communities, and cities. However, each dataset is usually a static snapshot from a specific domain, such as mobility, housing, or energy. Aggregating the data is highly desirable, but challenging because the datasets are usually ‘messy’ with different formats and quality.  



We aim to build a novel, AI-based, socio-technical system (AI system) to automatically put these snapshots together to form a comprehensive picture of the city. The data will be used to evaluate, predict and guide future city development. Integrating diverse datasets together can mitigate bias and uncertainty. Equal data access is a priority, so anybody can ask a question and be presented with information most relevant to them. We will use Austin as a prototype and then scale to other cities.  


We will: 

  1. Define the meaning of “equal data access” with our community and city partners through ongoing community engagement with resident stakeholders. 

  1. Design a novel information model to fuse different, ‘messy’ datasets throughout the city.  

  1. Build AI models to extract useful information for the public and mitigate errors or biases embedded in datasets, then predict impacts of urban development scenarios using historical data. 

  1. Evaluate the AI system from the perspective of ethical and technical experts, and with community feedback. 

Phase I involves data collection, domain mapping, and building the AI system with web access. 

Phase II applies the AI-based knowledge system to the first use case. The team will form an advisory board and analyze how Project Connect, a major Austin infrastructure addition, impacts transit, gentrification, housing, economic development, and more. 

Phase III will test additional use cases, which will be major infrastructure or business development projects, and assess associated impacts.  

project overview

This system’s utility will be wide-ranging and serve as a resource to industry, city officials, and public policymakers, as well as individual community members who are curious about how infrastructure development might improve or adversely affect their quality of life. 


Team Members
Civil, Architectural and Environmental Engineering
Arya Fahari
Statistics and Data Science
Devrim Ikizler
Dev Niyogi
Geological Sciences and Civil, Architectural, and Environmental Engineering
Andrew Waxman
Lyndon B. Johnson School of Public Affairs
Mingyuan Zhou
Information, Risk and Operations Management
Electrical and Computer Engineering