Rediscovering Cities through AI | Momin Uppal | TEDxSICAS DHA Youth

Quick Overview

Momin Uppal discusses how his team at the Center for Urban Informatics, Technology and Policy (CUI-TTP) at LUMS is leveraging AI and data science to redesign urban systems in Pakistan, specifically highlighting the creation of high-resolution population density maps for Lahore and optimizing emergency service routes by analyzing accident hotspots.

Key Points: The speaker, Momin Uppal from CUI-TTP at LUMS, advocates for evidence-based planning and policy design for sustainable urban systems using 21st-century technology like AI and data science. Lahore's built-up area increased four times between 1988 and 2021, leading to significant urban problems stemming from unmanaged growth. The team created high-resolution population density maps for Lahore by using an AI model for building detection on satellite imagery to disaggregate census data. The analysis showed that only 15% of Lahore's area (4.42 million people, 40%) is within a 5-minute walk of a public transport station. In collaboration with Punjab Emergency Services (Rescue 122), they analyzed accident hotspots using advanced analysis to improve resource allocation and response times. The team also worked with the Lahore Waste Management Company to optimize trash collection routes, potentially saving 150,000 liters of fuel monthly. The core research team includes urban planners, data analysts, engineers, and transport planners.

Context: The presentation, delivered at TEDxSICAS DHA, focuses on the work of Momin Uppal and his multidisciplinary team at LUMS' Center for Urban Informatics, Technology and Policy (CUI-TTP). The core theme revolves around utilizing modern data science and Artificial Intelligence (AI) techniques to analyze and subsequently redesign critical urban systems in Pakistani cities, using Lahore as the primary case study for population density mapping and emergency service optimization.

Detailed Analysis

Momin Uppal introduces the challenge of redesigning urban systems by first establishing the context of rapid, unplanned urban growth, citing that Lahore's built-up area increased fourfold between 1988 and 2021, resulting in significant environmental, health, and mobility crises. He emphasizes that any meaningful urban change requires high-resolution data, such as population density maps, which his team developed using an AI model for building detection on satellite imagery. This process allowed them to disaggregate census data into high-resolution formats. A critical finding shared was that only 15% of Lahore's area (covering 4.42 million people, or 40% of the population) has access to public transport within a five-minute walk. He also detailed a use case with Punjab Emergency Services (Rescue 122) to analyze accident hotspots, aiming to improve resource allocation and response. Furthermore, they collaborated with the Lahore Waste Management Company to optimize trash collection routes using AI, projecting fuel savings of 150,000 liters per month. The presentation concludes by showcasing the multidisciplinary nature of their team (urban planners, data scientists, engineers, policymakers) and promoting their ongoing work through the City Atlas project.