Dreamt up in sixties California, that vision now seems a long way off. But algorithms and artificial intelligence (AI) already have considerable oversight of our day-to-day lives – and recent research shows that they are now playing an important role shaping our environment.
Capable of analysing vast amounts of data much quicker than humans can, AI is already influencing the design of cities and infrastructure, according to a recent survey by London construction engineering company Arup. The research found that more than a third (36%) of architects, engineers working in the built environment and city planners rely on AI daily, with more than 80% using it at least once a week.
The report, Embracing AI: Reshaping Today’s Cities and Built Environment, includes responses from 5,000 professionals working in 10 countries: the UK, Australia, Brazil, China, Germany, India, Indonesia, Nigeria, Singapore and the US. They reported using advanced tools such as large-scale simulations and evolutionary algorithms to tackle complex work, including project design, advanced modelling and urban planning.
“AI's role in city planning and construction is expanding rapidly,” says Matthew Chang, co-founder and principal engineer at Chang Robotics in Jacksonville, Florida, who was not involved in the research. Applications include tracking mobility data, controlling traffic signals and developing predictive models for transit planning, he says.
Arup’s own tools include UHeat, which analyses satellite imagery and climate data to provide recommendations for ‘nature-based interventions’ that limit hot temperatures without additional building work. Other programs include Autodesk Forma, which is being developed to unify workflows across teams that design, build, and operate the built environment, including automatic evaluation of environmental factors such as wind speeds, sunlight exposure and noise levels.
Urban projects are not the only ones that can benefit from AI tools. Partners working on HS2 used tools from American firm Alice Technologies to optimise construction sequences, for example, reportedly accelerating schedules on the Copthall Green Tunnel by three months and saving £2m on works at Euston station.
Putting people first
Deployment of AI-based tools is already well underway in planning and construction – but with so many people affected by each decision, ethical issues must be front and centre. Planners cannot simply take results from opaque systems at face value.
Ethical responsibility is a key focus for a group of researchers led by Aston University in Birmingham, which aims to provide a way for urban mobility planners to create greener cities using evolutionary AI.
“The ethos of what I do pretty much has people in the centre,” says Dr Alina Patelli, senior lecturer in applied AI at Aston, who co-leads the group with Dr Dalila Ribaudo, senior lecturer in international business. “There is a known paradigm that's called ‘human in the loop’, but what I do is more than that – it has people at its core,” she says.
“Ultimately, what we do in research, engineering and academia is serve the people, societal good… It's quite a demonstration of hubris when people, some of my peers unfortunately, consider that they know better.”
Also involving experts from University College London, Ruralis University in Norway, the University of Turin in Italy and the Lisbon University Institute, the team has received £10,000 in funding from the British Academy. It plans to apply for EU funding to develop the platform, which would offer decision-makers urban mobility insights. Output would be based on data from city-wide sensors that is “crunched” in the AI’s core, providing policy recommendations in natural, easy-to-understand language.
Unlike mainstream AI tools such as large language models, Dr Patelli says evolutionary AI (inspired by the way biological organisms evolve) is “intrinsically explainable”, providing an explanation of its work alongside the developed model and its recommendations.
She gives the theoretical example of Transport for West Midlands wanting to redraw the Birmingham clean air zone in response to commercial concerns from local small businesses. The platform would consider data as varied as business revenue, air quality, traffic movements and ‘anomalies’ such as large sporting events or party conferences. It would then produce a model that describes everything, predict potential changes to mobility in the city centre, then generate recommendations, with justifications included. The output would include maps as well as text.
Bot the builder
According to the Arup survey, the “vast majority” of respondents (91%) believe it is important to have ethical guidelines for AI in the built environment. Roughly half (49%) also report apprehension about the dominance of global tech companies in AI development.
Overall, however, the report describes feedback as “overwhelmingly positive”, with almost two-thirds (61%) seeing AI as an opportunity. Only 11% see it as a risk to jobs.
AI will have a “complementary role” in future, Dr Patelli predicts. “It will be there as a tool to responsibly automate the mechanical, monotonous, soul-draining parts of our day-to-day jobs, including the activity of decision makers, such that they can draw insight from terabytes of data they can't possibly process in real time.”
Chang expects urban planning to include “AI-powered” architecture, design and engineering in the medium to long term. “For urban construction, we'll see municipalities employing robotic concrete systems, robotic fabrication,” he adds. “AI will play a strong role in research projects to progressively improve sustainability in the built environment.”
AI use in the sector will likely accelerate, says Wyatt Mayham, CEO and co-founder at Northwest AI Consulting. “If Arup does another survey next year I would not be surprised if that 36% number doubles, because there are so many interesting use cases.”
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.