Which industries have embraced digital transformation?
The advent of advanced industrial activity has led to the prospect of machines taking over more menial jobs, improving productivity, bolstering safety and generally making our lives easier.
In the past few decades automation has transformed a number of sectors, notably automotive and other manufacturing industries such as plastics processing.
While some tasks have become far more efficient through automation, jobs have been inevitably affected. Some studies suggest that nearly 2 million jobs have disappeared in UK manufacturing since 2000 due to robotic technology, for example.
What are AI and machine learning?
Robots are one thing, but in recent years other forms of intuitive technology have made inroads into the world of work, notably artificial intelligence (AI) and machine learning.
Many people are familiar with both concepts. AI simulates the processes of human intelligence in machines, notably computer systems and, for some, focuses on three cognitive skills: learning, reasoning and self-correction.
Machine learning works in two ways: supervised learning, where a machine is trained by using existing or known input and output data so it can predict future outputs; and unsupervised learning, where a machine finds patterns or structures hidden in data, which are then used to draw conclusions and take action.
Technologies such as AI in construction are a relatively recent addition to companies’ armory. It is well known that the sector is keen to find ways to improve productivity and glean greater efficiencies through using new techniques and processes.
Benefits of AI and machine learning in construction
Streamlining project delivery
One financial incentive is that large-scale projects are particularly at risk from being delivered late; according to consultant McKinsey, 98% of megaprojects suffer cost overruns of more than 30%, and more than three-quarters (77%) are at least 40% late.
The role of information is increasingly important for construction as it attempts to tackle such issues, but this brings its own set of challenges. On any job, but particularly on a large project, contractors will find themselves working with huge amounts of data every day, ranging from design changes to material availability, worker shortages and delivery issues.
AI and machine learning in construction give contractors and other stakeholders the opportunity to predict with a degree of certainty what needs to be addressed and to what level of urgency.
Improving worker safety
Such technology can also assist those in charge of making sure that a construction site is kept safe for those working on it. AI helps contractors by assessing risks and helping mitigate them, either through training or identifying where a particular risk might lie.
Using AI and machine learning for risk management sounds like a no-brainer, and in the UK, where the government has thrown its weight behind becoming a ‘technology superpower’ by 2030, there is considerable interest in taking up the opportunities waiting for industries like construction.
The future of AI and machine learning in construction
According to a review by professor Dame Wendy Hall and Jérôme Pesenti, ‘Growing The Artificial Intelligence Industry In The UK’, AI could add an additional US$814 billion (AU$1,220 billion) to the UK economy by 2035.
Given AI and machine learning’s reliance on data, access will need to increase, the review argues. With this in mind, the report calls for the creation of data trusts ‘to improve trust and ease around sharing data’.
Data protection concerns aside, and much as it could be a potential godsend to the sector, the adoption of AI and machine learning in construction is not going to be without its challenges.
Challenges of implementation – and then some
As the MIT Sloan School of Management points out: ‘Machines are trained by humans, and human biases can be incorporated into algorithms – if biased information, or data that reflects existing inequities, is fed to a machine learning program, the program will learn to replicate it and perpetuate forms of discrimination.’
MIT highlights the issue of algorithms that adopt offensive content, such as racist tweets and other social media content. It can effectively be summed up as: ‘Rubbish in, rubbish out.’
One study, ‘Opportunities and Adoption Challenges of AI in the Construction Industry: A PRISMA Review’, published last year, highlighted the issue for construction.
‘The construction industry should take advantage of AI throughout the project lifecycle, as it brings significant advantages’, the review says. ‘The increasing complexity of modern construction in pre-construction, procurement and post-construction stages is now the main driver for developing interest in digital technologies.
However, AI applications for the industry are only in the initial stage, and there remain significant research gaps that turn digital inventions for construction sites into reality. The construction industry needs to close this gap by looking at how AI applications are being used in other industries.’
The review further argues that the landscape of AI in construction is consistently changing, and the opportunities will likely far outweigh the challenges in the long run, once AI technologies mature.
‘Since AI needs a considerable amount of data for algorithm training, large-scale companies are likely to be more advantageous in the short run. However, these benefits will spread to medium and small-scale companies shortly when they recognise the cost and time benefits it may bring’, it adds.
Modern methods of construction are increasingly becoming a part of construction’s offering – think modular techniques – so it isn’t far-fetched to think that more advanced technologies will become part and parcel of the sector in the not-too-distant future.
However, practitioners and regulators will be mindful of the need to oversee such practical assistance in a pragmatic and effective way.