Risk and reward: managing the rise of AI in industrial automation

Introduction

Over the past year, artificial intelligence (AI) has rocketed into the spotlight, triggering both anticipation and concerns. It has swiftly transitioned into our everyday reality, bringing forth numerous practical applications.

AI’s transformational impact is most conspicuous in the realm of knowledge-based industries, where it has been revolutionising tasks associated with language, images, and data. I’m referring to generative AI, which goes further than simple rule-based measures and uses patterns and structures in data to generate ‘newish’ content and rules. But despite the excitement around recent developments, AI isn’t independently making the physical products people need to live their everyday lives. Instead, it is industrial automation that is driving accelerated production and transportation of these physical items. It’s more about delivering concrete value to the world than simulating intelligence.

The progression started over a decade ago with Industry 4.0, representing the 4^th industrial revolution resulting from the digitalisation of industrial processes. Industrial automation is surging ahead now because of recent worldwide supply chain chaos and labour shortages, due in no small part to the COVID-19 pandemic. We are therefore entering a new era, which promises monumental growth, yet doesn’t come without its challenges, including managing how AI will be implemented to help achieve efficiency, growth, and quality.

Expect the unexpected

AI has already been used to develop automated processes in several ways, from predicting customer demand to robots learning new tasks and overcoming obstacles with agility, spatial awareness, and pattern recognition on the factory floor. It has yet to become clear, however, exactly how generative AI will continue to improve these abilities. Every groundbreaking leap in technology brings forth a fresh set of trials and tribulations. As businesses embark on their journey beyond Industry 4.0, let’s call it Industry X.Y, they must prepare for a number of possible risks associated with the further implementation of continued advancements in AI in automation.

While the highly automated facilities of the future promise operational efficiency, disruptions will still arise, just in different ways. New, sophisticated technologies will therefore demand problem-solving workers equipped with multifaceted skillsets. This will no doubt include knowledge and skills in areas such as coding, robotics, manufacturing technology, and electromechanical hardware.

It is vital that, when a robot encounters a malfunction, staff with relevant expertise can promptly diagnose and rectify the issue. This means the substantial costs of business interruption, which may continue to increase into the future, can be mitigated.

As machines get smarter, the danger of cybercrime also increases. With fewer people on factory floors, there is a heightened threat of cyber attacks or other incidents destroying physical machinery before anybody has time to respond. Industrial control systems, for example, provide potential routes for attackers to cause physical damage, via online connections, to automated systems if not appropriately protected. Outside control over something as simple as air conditioning units, let alone autonomous devices on production lines, presents a significant risk.

While intelligent sensors may identify these problems, the effectiveness of automated responses hinges on uninterrupted functionality, including during fires, floods, or power outages – and this can be more challenging, with greater potential adverse consequences, than ever before.

Riding the automation rollercoaster: mitigating the AI-related risks

There are multiple ways for businesses to prepare for this new era of AI (and future generative AI) enabled industrial automation. Think of it as a rollercoaster ride: exciting, but with unexpected twists and turns. A well-executed transformation plan, that accounts for scarce skilled labour and the transition and/or integration of legacy equipment, can help to ensure a smooth ride regardless of which stage of the Industry X.Y. technological revolution a business is in.

As part of any plan, emergency response procedures should be developed to fit the expected increase in reliance on technology and reduction in human presence. Adapting emergency protocols in this way means they can accommodate for shifts in the work environment and lay out new ways for coping with interruptions. These often follow a revision in process technology. Due to the greater consequences of failure, changes in response plans have to be in lock step with changes in technology.

A greater prevalence of AI-backed automation technology in manufacturing industries, requiring different maintenance and troubleshooting techniques, means businesses should be prepared to invest in monitoring equipment and staff training. After all, well-trained staff will become the guardians of smooth operations and condition-based maintenance.

Prioritising change management by recognising the pivotal role of software in future business operations can also help navigate some of the expected challenges that come with increased automation. Leveraging digital twins is a good way to test new system design, software versions, and other changes in advance to avoid physical failure.

Creating or partnering with innovation labs or technology incubators could also be an option for businesses preparing for AI-powered automation. They act, essentially, as a testing ground for the latest and greatest innovations, ensuring they won’t disrupt business operations when they go live.

Looking forward to the benefits of AI-backed automation

When the risks are carefully considered and mitigated against, a company can concentrate more time and investment on the positives.

Firstly, by changing the role and factory presence of human workers, automation can reduce exposure to workplace risks and simplify the management of a diverse workforce. This is expected to lead to fewer errors, provided that the automated systems are well-designed and well-maintained.

Even with future developments to AI in automation processes, and the greater abilities that might bring for automated systems, human collaboration with robots will remain essential. While AI-powered robots might generally assume more tasks, particularly the most tedious, it means humans are able to focus on more complex roles. This transformation is all about filling vacant positions and improving the efficiency of the roles performed by human workers.

Secondly, automation contributes to environmental sustainability. Reduced building occupancy and a more efficient physical footprint result in lower energy consumption for heating and cooling, with significant implications for global electricity consumption.

The potential strides to be made from AI-backed improvements to automation offer undeniable promise, meaning businesses across various sectors are diligently exploring ways to capitalise on the benefits of automation. This means that now is a prime opportunity to address potential challenges proactively, ensuring a smoother transition and cost-effective issue management.