Comment: Engineers must adapt to AI or fall behind

Artificial Intelligence (AI) technology is simultaneously the most boring and the most exciting development of our generation.

It’s boring because beneath the hype lies the unglamorous reality of needing to sort out our data. But it’s exciting because of all the potential applications if we get the foundations right.  It can help us transform how we design, construct, and maintain our ever-changing world.

MORE FROM OPINION

The fundamental takeaway is: it’s here to stay, so adapt now or risk falling behind.  

While most people are becoming increasingly familiar with generative AI – the technology behind chat bots and AI-generated images – this only scratches the surface of the wider field of artificial intelligence. There are many other forms of AI that are already transforming our industry for the better, for example:

·         Computer Vision is a field of AI that teaches computers to understand visual information from the real world; it can be used in engineering for data collection and collation. For example, it can capture information from photos and video footage, generating insights to track work progress and productivity in real time. It can also help to identify potential risks by analysing the degradation of operational assets over time.

·        AI is already being used to enhance planning and project management by helping to analyse project timelines and dependencies, and to optimise resource allocation.  It’s a clear route to reducing costs and delays.

     Digital twins can use Machine Learning in a number of different ways to help improve decision-making.  And one of the most exciting developments is to replace the physics-based models at the heart of some digital twins with AI-based models that are quicker and cheaper.  

·       This year’s next frontier – Agentic AI – offers the possibility of joining each of these different versions of artificial intelligence, though not without raising important questions about oversight and control. Unlike ‘traditional’ AI systems that require human intervention for decision-making, Agentic AI can operate autonomously and independently, completing objectives with little or no human supervision. Whether its deployment in our industry will succeed or not, we will have to wait and see.

However, despite these clear opportunities to identify risks, reduce costs, and improve efficiency, there is very inconsistent uptake of AI tools among engineers. If every organisation has a completely different approach to AI, it will create a disjointed mix that simply won’t benefit us.

The inconsistency is due to the fragmented nature of our industry. In such a broad sector, encompassing so many separate organisations, it’s difficult for us to establish a joined-up approach, even if it would benefit us all. To be successful, we must have industry-wide standards and a joined-up approach, particularly when it comes to the data on which all of the promises of AI depend.

Back to the boring bit.  To get the full value from AI we need to get our data right.

Fundamentally, we must address the perennial issues of data quality, data access, and data security. Beyond that, we need to grapple with boring but important questions:

When designs can be completed in seconds, who pays whom for what?  When AI-enabled designs fail, who bears liability?  Do we change the business models? Do designers merge with insurers? These are amongst the pressing questions that the industry will need to answer.

In all of this, we need some kind of consistency to emerge.  Yes, every project is unique, and every client is different, but we will go further and faster if we do so together.  This is not an argument for a ‘one size fits all’ approach, but it is a call to identify good practice and share it in a coordinated way.  A little bit of coordination would go a long way – we need a shared strategy for effective AI adoption.

While this technology is moving at a rapid pace – we mustn’t lose sight of who’s in charge.  Like any digital transformation, we must make sure that people are at the centre of it.  There is always a danger that we get so excited about a new technology that we inadvertently get people to serve it, rather than the other way round. 

On the question as to whether AI will replace us? No, not if we use it properly, but AI will most certainly separate those who use it well, from those who don’t.

AI is with us and it’s here to stay; boring, exciting, threatening, and promising all at once. We just need to adapt.

Mark Enzer OBE FREng is an ICE specialist advisor and Mott MacDonald Fellow