Artificial intelligence (AI) is becoming increasingly more important to the management of critical national infrastructure as we search for better economic, social and environmental outcomes from limited public investment.
Much has been said about the emergence of AI as both a game changing tool to achieve greater efficiency and a relatively unknown threat to society. For all the debate, the reality for infrastructure professionals is that this technology is already here and has presented us with a unique opportunity to gain unparalleled levels of operational efficiency. Given its immense capability for providing operational efficiencies and expediting decision making processes, it is perhaps no surprise that AI is already at the forefront of the plethora of digital tools that infrastructure professionals so heavily rely on. Especially given its ability to combine with and augment the power of other digital technologies such as data analytics.
The combination of AI and data analytics has now accelerated data analysis even further as it can enable us to gain valuable insights and improve decision-making across the board at a rate and efficiency not encountered before. However, in the midst of all this, human oversight will remain crucial in not only in ensuring the efficient use of AI solutions but also in the responsible and accountable use of the technology in line with overarching priorities and principles.
As AI technology becomes more commonplace and essential to business operations, how can we avoid the potential pitfalls and unintended consequences that come with this powerful technology?
A case for AI adoption – smarter, slicker operations
With increasingly limited public funding, it is no surprise that infrastructure asset management providers are looking to embrace revolutionary new technologies like AI to streamline operations, enhance safety, and drive efficiency.
Predictive AI is one example of intelligent data analysis that is presenting clear opportunities for improving existing manual operations. For instance, it is fast becoming a mainstay feature in infrastructure asset maintenance systems. This includes its use in remote asset condition monitoring tools, geospatial and environment management systems as well as traffic and timetable management. As a result, we can now predict and track the wear and tear of assets, monitor vegetation growth on rail tracks using satellite or drone footage, and detect structural defects and damage using remote video capture sensors.
Case in point is Amey’s monitoring of the Forth Bridge, where the use of infrastructure monitoring and analytics systems leveraging AI technology were used to gain real-time understanding of the health and status of the bridge through data derived from sensors installed along the bridge structure. Data insights from the bridge structure would then dictate any necessary maintenance repair works.
Another powerful AI tool that is transforming infrastructure operations is AI computer vision technology. Using machine learning algorithms, the solution can receive and analyse data in the form of images and videos to provide real-time actionable insights, as well as building a clear and consistent view of the conditions of infrastructure assets.
Whilst it’s clear that AI comes with a whole host of desirable operational benefits, there is still a lack of understanding around the efficient, responsible, and ethical use of the technology. Many organisations are still getting to grips with the different types of AI models available and crucially, how each one affects different decision-making processes, and by extension, operational efficiency. However, many more are yet to understand the potential downsides of AI adoption, particularly regarding concerns over privacy, security, ethics, and bias. These elements of AI usage can simply not be overstated or ignored and will require a great deal of human intervention and steering. It is human expertise that will be at the forefront of how we can get the best out of AI technology.
Ethics and safety guidelines for trustworthy AI
To ensure the ethical and safe use of AI, organisations need to establish a set of guiding principles that ensure the technology is implemented and used in a responsible way. This could include having an AI policy in place that outlines key practices for adopting and using different types of AI solutions, and central to this will be human decision making and expertise. It is therefore critical that whatever AI ethics guidelines a company uses, that it has the appropriate controls in place to safeguard these outcomes and prevent autonomous decision-making without vital human intervention.
After all, AI should not be replacing people within a workforce, but rather augment their work output. What we are increasingly seeing as a requirement from our customers is keeping the ‘human-in-the-loop principle, which enables colleagues to use AI to undertake the more repetitive, manual tasks, and focus more of their time on the more strategic endeavours.
Human expertise and AI intelligence – a winning combination
Businesses will without a doubt require human guidance and expertise to ensure that they are operating AI technology in a way that yields the desired operational efficiency and goals. As such, businesses will need to take a goal-centric approach to AI adoption. The focus should be placed on developing and adopting AI tools that benefit clients and end users.
In the context of infrastructure, AI driven outcomes should be safe, robust and reliable, which is why it is important that we combine human expertise and decision-making with leading-edge technological innovations. As a leading provider of infrastructure solutions, we recognise that AI, with its capacity for data processing and predictive analytics, is a powerful tool. However, it is the human engineers who bring critical judgment, contextual understanding, and ethical considerations to the table. Human expertise is essential for interpreting AI-generated insights in the context of certain variables that algorithms might not fully account for. This includes environmental factors, regulatory requirements, and community impacts. Human expertise ensures that AI-driven decisions align with broader project goals and safety standards, minimising the risk of errors that could lead to costly or dangerous outcomes.
For instance, when using AI solutions for predictive maintenance of critical bridge infrastructure, human expertise is required to assess the potential AI data insights taking into account factors such as the bridge's historical performance, environmental conditions, and traffic patterns. They are also expected to determine the appropriate maintenance works, ensuring that the AI's recommendations are feasible and aligned with safety protocols and regulatory requirements.
In a similar vein, AI models can also help predict the environmental impact of new infrastructure projects, such as highway expansions, providing insights into carbon emissions, land usage, and biodiversity disruption. Engineering teams will be required to apply their understanding of local ecosystems and environmental regulations to adjust things such as project designs to mitigate any potential negative impacts, ensuring that the insights from the AI solutions are both practical and environmentally sound.
As we embrace AI technology to foster innovation and create new opportunities to exploit data, it’s important that we balance these advances with human-centred approaches. The synergy between AI and human expertise is key to delivering infrastructure solutions that are not only innovative and efficient but also safe, secure, and responsible. This balanced approach ensures that we harness the full potential of AI while upholding the highest standards of engineering excellence. It is now time the industry embraces a more responsible and human-centric approach to AI adoption, where we seek out intelligent tools that are backed by human expertise to complement their power, usage, limitations - and potential positive impact on society.