ARGOS Project Blog By Jeffrey Tong, MSc Student in Regional and Urban Planning at LSE | Yale-NUS
Artificial Intelligence (AI)-assisted systems and technologies have been prevalent for a number of years and present in many domains familiar to our everyday lives. Algorithms for instance determine how suggestions are featured on one’s social media feeds, while routes and prices for private hire vehicles are optimised with the use of AI. AI-assisted decision making is set to become increasingly prevalent and embedded within our lives in the years ahead.
While AI research has gained prominence in a plethora of areas such as e-commerce, financial systems and scientific analysis, its potential to address some of our most pressing challenges is only just beginning to be realised within practice.
The Importance of Cities
With cities contributing to 71-76% of energy-related carbon emissions and 1.6 billion people in cities potentially exposed to high temperatures by 2050, cities are both sites impacted by climate change and simultaneously sites of climate action.
The recent COP26 has reinforced the role of cities in accelerating action amidst the negotiation roadblocks and hurdles among national governments. Coinciding with the major conference, the UK government for instance announced a £27.5m Urban Climate Action Programme to support cities across Africa, Asia and Latin America to transition to net zero on Cities Day (11 November).
AI in Urban Planning
Given the importance of urban environments, the practice of urban planning is a critical field for the application of AI. AI-assisted systems hold great promise to enable the development of smart, sustainable and resilient urban infrastructure amidst a changing climate. Reinforcement learning can potentially aid in understanding how buildings consume energy and to optimise energy utilisation within building management systems, while artificial neural networks and cellular automata can be utilised in urban planning to simulate development patterns and model entire cities.
Exciting examples are emerging throughout practice. Cities and urban planners have begun working with AI to enable more effective planning amidst urban challenges.
In Delhi, a network of more than 7,500 CCTV cameras and automated traffic lights carry sensors to track real-time traffic conditions. The insights from the AI-based traffic systems allows transport planners to implement adjustments to alleviate congestion within the city.
In Singapore, 3D semantic modeling has enabled the development of a Digital Climate Urban Twin that allows planners, engineers and architects to visualise real-time information such as weather, traffic flow and energy use in buildings and streets and how these forces may impact a building. Robotic Process Automation and Natural Language Processing have been deployed within the Urban Redevelopment Authority to codify land use planning guidelines and more efficiently respond to public feedback.
What Next
AI within the field of urban planning can aid in the trajectories of cities towards net zero, and its application is expected to become more prevalent in the coming years.
Beyond climate change mitigation, emerging research and use cases point towards how AI can potentially enhance cities’ overall resilience and adaptation to climate change, and under multiple hazards. In my preliminary literature review done within the research project on AI-assisted Resilience Governance Systems (ARGOS), examples of the use of AI to manage compound risks, or risks associated with multi-hazards, have emerged.
More broadly, the momentum in research and application of AI to address the multitude of society’s most critical challenges is exciting for both researchers and practitioners alike, and is worth paying further attention to.
The writer is a research assistant with the ARGOS project and is currently a MSc student in Regional and Urban Planning Studies at the London School of Economics.