Artificial intelligence: the fifth industrial revolution

AI in the industrial sector offers substantial advantages, but it is not without its challenges.

In recent times it has been hard to avoid the discussion around artificial intelligence (AI) and its impact on our daily lives. By now most of us will also have experienced generative AI (GenAI) in some form — the most well-known example being OpenAI’s ChatGPT — and have been maybe surprised by its abilities or perhaps underwhelmed. But these tools are general in application, and may not be useful in many circumstances. However, AI that is focused, with specific relevant data to learn from, has shown it can be truly useful.

The industrial face of AI has been under development for some years — as an important aspect of the Fourth Industrial Revolution — but we are only now beginning to see it become more than just a tool for only the very large companies.

Until now, only very large, technology-focused companies have been able to make good on the promise of Industry 4.0. They are part of what the World Economic Forum calls its Global Lighthouse Network¹ that showcases how digital transformation and cutting-edge technologies like machine learning and digital twins are able to achieve the automation dream. Among the 153 factories of the Network are various offices of companies like Siemens, GE Healthcare and Foxconn: companies that have the resources to focus on advanced automation.

According to McKinsey, there is a ‘chasm’ between the Lighthouse factories and everyone else in terms of technological maturity — they tend to be three to five years ahead of everyone else.² This was particularly evident after the COVID-19 pandemic: 85% of Lighthouse factories only saw a 10% or less reduction in revenue, while this was true of only 14% of others. Although they faced the same supply chain risks, 65% of Lighthouses were already dual-sourcing and increasing inventory by 2022, compared with only 24% of other companies.

In the last two years however, the readier availability of AI-based technologies is beginning to make it possible for smaller companies to begin to take advantage: lower-cost industrial IoT and associated cloud-based AI mean that a broader range of businesses can begin to realise some of the benefits.

Asset maintenance and process optimisation

Industrial businesses can now deploy low-cost sensor networks and take advantage of cloud-based applications incorporating AI to predict maintenance needs in advance, reducing wasted time and cost on unnecessary maintenance and preventing unforeseen downtime.

Where such companies are already using a manufacturing control system (SCADA or DCS) from an automation vendor, these systems are now being upgraded with AI-based tools to help optimise production. Many vendors are now also providing GenAI-based support tools for operators and technicians to help improve efficiency and learning. Such tools are not like ChatGPT: instead, they are using large language models (LLMs) specifically based on real and site-specific automation and process knowledge.

Improved robotics and automation

Current research in robotics is focused mainly on making robots work better. Industrial robots tend to be large and expensive, and very inflexible — they need to be manually reprogrammed when the process needs to be changed, and are difficult to coordinate to work together.

Robotics researchers are working to enable better robotic vision, coordination and flexibility through the use of AI and machine learning. This is taking time however; we might assume that since AI can now beat champions of chess or Go, that AI must be able to do everything well — but some of the simplest tasks for humans can often be enormously difficult for an AI-driven machine. We evolved for millions of years to be able to do the things we do and teaching an AI-driven robot to do these things has proven to be a difficult problem to solve.

Sustainability efforts

AI can significantly enhance environmental sustainability efforts by helping to optimise resource use, reduce waste and improve efficiency in various sectors.

The chief benefit of AI is in being able to analyse large amounts of disparate data quickly. AI systems can therefore be used to optimise energy usage in buildings, factories and even entire cities. The idea of a ‘smart grid’ has been discussed for many years, but AI will enable easier balancing of electricity demand and supply, reducing energy wastage. AI-based predictive models are being tested that can increase the reliability of renewables, integrating them better into the energy grid and reducing dependence on fossil fuels.

In manufacturing, AI can help minimise energy-intensive practices and cut down overall energy consumption, while in agriculture, AI can manage irrigation and distribution to ensure water conservation. By monitoring soil moisture levels, predicting rainfall and analysing water flow, AI can help reduce water usage and prevent resource over-extraction, especially in drought-prone areas.

In the recycling industry, AI-driven vision systems akin to those in advanced manufacturing plants can use computer vision to identify and sort recyclables and separate them from waste, improving recycling rates.

The other side of the environmental coin

Although AI can contribute to sustainability, the massive computing power that is being used to sustain AI systems has an energy demand and environmental footprint that must be carefully managed to ensure net positive impacts. Balancing these benefits and risks is essential as AI becomes increasingly integral to industrial processes.

According to Forbes³, “AI’s projected water usage could hit 6.6 billion m³ by 2027, signalling a need to tackle its water footprint.” The water need for cooling data centres due to the escalating demand for online services and GenAI systems has been estimated at 9 litres of water per kWh of energy used — a problem that should not be overlooked when assessing the use of AI for sustainability objectives.

Cybersecurity risks

The cybersecurity risks of industrial AI primarily stem from increased connectivity and reliance on digital infrastructure, although this migration to digitalised systems has been underway for some time.

What AI adds to the equation is its dependence on learning data. If an attacker can compromise the data that AI is being trained and updated from, it could be induced to make erroneous decisions, which could be catastrophic in a number of ways if the AI is heavily relied upon in the future. It is therefore necessary for cybersecurity protections to apply to the data itself, and raises questions about how and where an organisation stores and processes the data.

Meanwhile, cybercriminals are already utilising AI in their attack methodologies, but cybersecurity defenders are also using AI to speed up their response to incidents: in the future it may become an AI vs AI battle.

What about the social impact?

The fear that AI will lead to job losses is understandable, especially as automation advances rapidly. Many routine tasks across various industries are increasingly automated, which has already led to shifts in certain job types. However, AI’s impact on jobs is complex and multifaceted, leading to both displacement in some areas and new opportunities in others.

Eliminating routine and low-skilled work

AI and automation often replace repetitive and manual tasks, especially in manufacturing, logistics and data processing. For example, robots in manufacturing and automated sorting in logistics have reduced the need for human labour in these specific roles, but at the same time have reduced injury risk by eliminating heavy physical work. However, in fields like finance, automated processes handle data entry and simple accounting tasks, reducing demand for these types of jobs with no specific human benefit.

The positive side

AI is creating demand for new types of roles, especially those related to designing, programming, maintaining and improving the AI systems themselves. Positions such as data scientists, machine learning engineers, cybersecurity professionals, AI ethicists and robotics technicians have grown as AI adoption increases. Many AI-related jobs require specialised skills in programming, data analysis and machine learning; this shift can pose a challenge for workers without these skills, but it also opens up opportunities for people willing to retrain or upskill.

AI can also serve as a tool to augment rather than replace human capabilities. For example, AI can helps automate data analysis or generate reports from structured data, enabling human workers to focus on more complex work, and avoid tedious activities.

A social as well as industrial revolution

As McKinsey² put it: “What steam was to the first Industrial Revolution is what AI will be to the fourth. And much as coal supply chains and factory infrastructure were the tipping point that enabled steam power to race up the adoption curve, data collection and data infrastructure are doing the same in the fourth.”

AI in the industrial sector offers substantial advantages, particularly in enhancing productivity, quality control and operational efficiency. By automating repetitive tasks and enabling predictive maintenance, AI reduces downtime and optimises the use of resources, helping to lower costs, improve environmental sustainability and minimise waste.

However, the rapid adoption of AI also brings challenges. Job displacement is a concern, as automation replaces certain manual roles, creating a need for retraining and upskilling. Additionally, implementing AI can require significant initial investments, and the technology introduces cybersecurity risks, as highly connected systems become more vulnerable to cyberthreats, and while AI can contribute to sustainability, its energy and water demands are very high.

Balancing these benefits and risks is essential as AI becomes increasingly integral to industrial processes.

<sup>1. World Economic Forum 2024, Global Lighthouse Network, <<https://initiatives.weforum.org/global-lighthouse-network/home>>
2. McKinsey & Company 2024, Adopting AI at speed and scale: The 4IR push to stay competitive, <<https://www.mckinsey.com/capabilities/operations/our-insights/adopting-ai-at-speed-and-scale-the-4ir-push-to-stay-competitive>>
3. Forbes 2024, AI Is Accelerating the Loss of Our Scarcest Natural Resource: Water, <<https://www.forbes.com/sites/cindygordon/2024/02/25/ai-is-accelerating-the-loss-of-our-scarcest-natural-resource-water/>></sup>

Image credit: iStock.com/tanit boonruen