Editorial

Artificial intelligence (AI), smart cities, and new innovations in science and technology create new types of megaprojects and can fundamentally alter the management, design, and delivery of existing ones. Historically, megaprojects built the cornerstones for massive urban, regional and sometimes national transformations signaling growth and wealth in the ages of neoliberalism and globalization and before. It is contended here that technological advancements of this kind are slated to follow a similar path with an important distinction, sustainability must be achieved as a determinant to our life on earth with AI evolving as the potential key to effectively address the impending climate crisis (Nishant, Kennedy, and Corbett 2020).

In this special issue, contributing authors are exploring the role of infrastructure megaprojects and their contribution to sustainable development during this transition towards our future with AI. Megaprojects, in this sense are analyzed in a dynamic, complex, and connected way. They are examined jointly as spatial infrastructure developments that are planned, appraised, designed, and constructed to contribute to the significant formation of places and as intelligent machines programmed to continuously learn to optimize the better use of this space. The contributions critically examine theoretical and practical aspects at the intersection of decision-making concerning infrastructure megaprojects, AI, and sustainability that crucially shape our future. Collectively, the authors of the special issue help clarify concepts of AI, exemplifying AI as an evolving tool that has the potential to govern infrastructure megaprojects to drive ecological, environmental, and socio-economic changes for the better.

The theoretical contributions in Part One of this special issue look to conceptualize artificially intelligent infrastructure megaprojects and set boundaries to existing definitions by creating a new understanding concerning the role of autonomy in decision-making brought about by these changes. By drawing on lessons from concurrent governance (Waltz and Firth-Butterfield 2019), they critically question when, how, and why AI can be an enabler of institutional change for power (re-)distribution, civic engagement, and bias removal. Advancing current knowledge (Kassens-Noor et al. 2021), they highlight in which ways AI has the power to significantly shift societal, political, and environmental agendas and values.

In Part Two of the publication, the authors also examine evolving artificially intelligent megaprojects as case studies of impending challenges and opportunities to foster environmental and social health given economic pressures. Aligned with Dimitriou, Ward, and Wright (2017), the authors collectively reject the idea of the ‘iron triangle’ project management approach commonly used to judge project ‘success’ because continuous learning - driven by new IT developments, advances in communication technologies, and new efficiencies in historical Big Data analyses – can constantly redefine what are considered to be ‘good’ outcomes for different places. From this perspective, the special issue illustrates the conditions under which AI is sustainably deployed in megaproject planning, appraisal, design, and delivery to enhance climate resilience. The practical and theoretical contributions shed light on the ways AI is jointly driven by private companies in search for ever-greater efficiency by governments in their aspirations towards sustainability, and by citizens across the world who adopt, alter, or reject new instantiations of AI within their everyday lives. Much of the concurrent discussion of AI and infrastructures remains speculative, and the authors urge the readers not to lose sight of potential challenges that are not yet clear in practice.

The first paper of the special issue is written as a collaboration between academics and practitioners at the cusp of new innovations in technology. Eva Kassens-Noor et al jointly propose a new research field called ‘scAInce’ that conceptualizes how and why AI influences and impacts our world as we know it. In their paper, they define AI megaprojects as a synergistic and enduring merger of AI software for large-scale infrastructure planning, appraisal, design, construction, and management of delivery. The authors conceptualize how AI megaprojects as virtual AI futures might enable sustainability in our built environment.

In the second paper, Hintze and Dunn differentiate how AI can be directed towards public or private interests and argue that cases in which AI agents act autonomously without direct human intervention present novel questions and challenges about how to achieve sustainable outcomes. They show how the use of AI for decision support systems, the programming of objective functions for AI that acts autonomously, and the regulation of AI’s unanticipated behavior, each present different opportunities and challenges for regulation. The authors contend that an important distinction among these approaches is often unclear in much of the existing literature, calling instead for AI to be delivered in the public interest. They illustrate their position in the context of cases in smart electricity grids and autonomous vehicle intersection coordination.

In the third paper, Linke et al lay out a vision for a sustainable, climate-resilient, and artificially intelligent megaproject, citing the ‘eHighway’ (electric highway) as an example. Referred to as the ELISA Project, five prototype trucks here run on the largest e-Highway test track in the world in Germany. The authors analyze the strengths, weaknesses, opportunities and threats for AI in the eHighway system as a sustainable mobility response for long-distance road freight transport. Triangulating data, energy, and technical availability of this system, they found that during the pilot phase new unpredictable incidents occur frequently which would be hard to detect by AI. Long-term, however, the authors argue that humans will no longer be capable to process the wealth of data and AI will inevitably become necessary. This paper provides the first glimpse into the possibilities of and challenges to applying AI in an eHighway system.

In the fourth paper, Decaminada contributes to our understanding of Hyperloop technologies that comprise tubes, pods, and terminals to form a novel high-speed transportation system. The author analyzes public perceptions and people’s expectations of such visionary technologies. Using text-mining from the contents of tweets, the author finds the public at large has already formed unrealistic expectations regarding the technology’s potential use. He argues that twitter users overstate the technologies’ potential benefits, while understating the potential costs of construction. The author further found that about half of the Tweeters hold unrealistic expectations over the speed of implementation of that technology. Thus, he concludes, caution is necessary in spending public money driven by mere advocacy championing an unproven technology.

In the fifth paper, Lehmacher et al examine the latest developments in AI technologies to reduce port congestion. Ports are critical mega infrastructure projects where traffic flows need to be managed well (Dimitriou, Ward, and Wright 2017). Using the case study of the Port of Valencia, the authors describe a new methodology using AI to enable the optimisation of traffic flows. The authors report that early results of this application have shown to be robust and reliable, while the impact on traffic flows has provided enhanced efficiencies, eased congestion, and reduced greenhouse gas emissions. They forecast that AI-powered implementation of this approach can not only benefit port cities but entire ecosystems such as the global maritime supply chain network.

Around the world, governments are piloting AI technologies to develop their regions and cities, and increasingly, national development and security (Allam and Dhunny 2019; Cugurullo 2020; Macrorie, Marvin, and While 2021; Trencher 2019; Yigitcanlar et al. 2020). Megaprojects are increasing in scale, scope, and size and require sustainability and resilience to be successful. The overarching message of this collection of papers is that for AI to be useful in managing the complexity of decision-making for megaproject developments in all sectors, we must carefully scrutinize and evaluate the assumptions and the objectives underlying the AI decision tools use, particularly as regards to how they serve to promote and inform sustainability. To achieve this with any sense of integrity for AI to support intelligent and sustainable mobility solutions, it must be supplied with reliable data, transparent, carefully governed, and provide realistic accurate expectations of what can and cannot be achieved.

Eva Kassens-Noor
Institute of Transport Planning and Traffic Engineering, Technical University of Darmstadt, Germany
[email protected]