The Fourth Industrial Revolution (4IR) is already in full swing, with emerging technologies that are transforming our world. Entire chains of production, management and governance are being disrupted, and nowhere is innovation having a bigger impact than manufacturing. As a result of Artificial Intelligence (AI), increased automation, Internet of Things (IoT) and robotics, industrial operations have become both smarter and faster.

These technological advances are also helping manufacturers save valuable time and capital by rendering processes more efficient and economical, creating unprecedented possibilities for growth. This is why the 4IR is expected to generate AED 13.6 trillion in global manufacturing. A McKinsey Global Institute simulation revealed that those who adopt AI will experience a 122% increase in their cash flow.

 

Importance of Safety and Security for Humans

 

At the core of the 4IR is the transfer of decision-making from humans to machines. This raises important questions about the safety and security of humans who are working with machines – sometimes in such close proximity that they are called collaborative robots, or “cobots”.

 

One potential issue lies in Machine Learning (ML), which is the process of getting machines to learn how to behave through repetition. However, the reality is that a machine cannot possibly learn how to perform a task in all situations and may not adapt appropriately when a new one arises. This is one way the 4IR creates uncertainty and possible risk to humans.

 

Another potential problem is “reward hacking”, where the robot or cobot learns some desirable behaviour but ignores others. For example, the machine can learn to move quickly from point A to B. But can it learn to stop upon encountering something or someone unexpected in its path? And if so, can it do so quickly enough to avoid impact and injury? Since not all potential scenarios can be anticipated, weaknesses in ML may not show up in safety testing.

 

Other challenges arise whenever the production process is revamped, and automation must adapt, or in the case of software updates, which can also result in changes that can expose ML weaknesses.

 

Keeping humans who work closely with machines safe and secure is just as important to consider as the efficiency and cost-savings they generate. To achieve this goal, industrial enterprises must tackle the unique challenges brought about by the 4IR and guarantee safety assurances. This includes clearly defining a system’s behaviour under all circumstances, adjusting that behaviour accordingly and providing education and training on potential issues that may arise – and acting on them.

 

  

Challenges and Potential Solutions in 4IR Manufacturing

 

As outlined in a briefing paper commissioned to Policy Links at the Institute for Manufacturing, the University of Cambridge by the Global Manufacturing and Industrialisation Summit (GMIS) and Lloyd’s Register Foundation (LRF) on ‘Safety Assurance of Autonomy to Support the Fourth Industrial Revolution’, cobots can be viewed as an evolutionary step in factory automation. Yet it is much harder to predict the safety of an autonomous machine at work compared to automated systems.

 

Within each and every factory site, human reasoning must come into play to ensure a safe outcome. These include judgement, experience, instinct, skills and contextual awareness. Furthermore, to tackle the safety challenges in a situation where the machines move about and interact with humans, added foresight, care and consideration must be taken.

 

At this point in time, the 4IR is inevitable. And as the field addresses concerns over safety and security, the benefits remain compelling. In the coming years, greater collaboration between the manufacturers, industry, researchers and governments is essential to explore the full potential of the 4IR. An example is the United Arab Emirates Fourth Industrial Revolution Strategy, which was implemented to create policies and regulations to address the challenges of the 4IR while ensuring the wellbeing of citizens. By creating this strategy, the UAE has reinforced its position at the forefront of adopting cutting-edge technologies that can generate previously unimagined efficiency and sustainability – as well as safety – across society as a whole.

 

Moving forward, continuous monitoring of a machine’s behaviour is crucial to keep the humans who work alongside them safe. Importantly, human workers must be empowered with adequate training to act early and respond rapidly.