Industry 4.0 is a hot topic, but this term – which is more frequently used as a buzzword than to convey any specific topic – is hard to unpack. Usually it is used to convey a somewhat vague and undefined concept of the future of industry, with factories equipped with cutting-edge technology, but all of this is rather unspecific and fails to tell us what exactly industry 4.0 will look like, what it means for manufacturers, and how it will come to be.
Understanding Industry 4.0
Industry 4.0 is a German idea, with its origins among the future-thinking German engineers who want to see industry advance in line with technology. At its most basic level, Industry 4.0 means going beyond the automation of work as we’ve seen over the last couple of centuries and moving towards the computerisation of work and manufacturing.
Building on the trends of automation, Industry 4.0 will draw on information technologies in order to facilitate greater levels of data exchange in manufacturing technologies. It includes concepts and technologies like cyber-physical systems, the Internet of Things (IoT), and cloud computing. Astute techies will have probably heard of at least two of these, namely cloud computing and the IoT.
Cloud computing is an undeniable trend across all businesses. The death of the company-owned data centre is near at hand. Now, more and more companies are looking to the cloud to answer their physical infrastructure needs. Cloud service providers (CSPs) are increasingly competitive, which has aided in driving innovative infrastructure solutions. AWS, or Amazon Web Services, is perhaps the leader, with a wide range of offerings, including processing time for ready-to-go machine learning, which means any application can take full advantage of artificial intelligence. While this may make coders and techies swoon, what does it mean for industry?
The answer lies in information and data. We live in an age where the new oil is data, where we are now inundated in more information and data than ever before, and more data is created daily. Computer scientists and engineers are largely in agreement that humans are becoming less effective at analysing information and making use of it. Algorithms, however, can help us, and herein lies the need for machine learning. AI neural networks and machine learning can assist us in understanding large volumes of data. For the new industrial plant of the not-so-distant future, having off-the-shelf machine learning in cloud services not only means that there is a place, off-site, to store all the huge amounts of data being generated by plant processes, but this data is analysed in real time – informing the decisions of human plant operators, helping to reduce downtime of mission critical machinery and processes.
The potential for this is huge. Industry 4.0 will lead to the smart factory. Data will be generated within the modular structured smart factories by cyber-physical systems that monitor processes, creating virtual copies of the physical world and, using real-time collected data, are able to make decisions independent of human evaluation. Cyber physical systems will require IoT devices, which are increasingly popular in our day-to-day lives. Take smartphones interacting with thermostats, or digital watches that automatically sync to phones and laptops, collecting data on our movements, exercise routines, calorie loss, and more.
Using IoT, cyber-physical systems will speak to each other, which is a human way of understanding data flows and data sharing amongst devices. This will give humans the ability to see everything that is happening in the factory. But with so much data, AI systems will become increasingly necessary to interpret and make sense of that data, informing humans as to where problems in processes or machinery have been identified, where there are production losses or gains or, for example, where re-supply of industrial gases is required.
But industry 4.0 goes beyond the smart factory and the production line, it extends out from here and into the world beyond, interfacing with the entire value chain and the supply chain. In the future, data transfer and AI will enable producers and manufacturers to identify supply and demand by geographic locations with real-time accuracy, adjusting manufacturing requirements alongside this information, and preventing loss of money or energy wastage.
Going even further, this level of control will prevent unnecessary environmental damage, it will harness energy consumption, and bring manufacturing inline with the environmental needs of today’s world. It will usher in the end of run-away fordism, and enable manufacturing to be clean, manageable, and efficient.
A closer look at Industry 4.0
While we can paint a wonderful vision of the future using terms like AI, IoT, smart factories and cloud computing, we are a long way off from the clinical looking factories we see in such science fiction classics like Ghost in the Shell.
The move is still in its early stages and, most likely, there will be disruptive technologies that help to shape the future of Industry 4.0. But what are the main principles behind the design, and what are manufactures already doing to embrace the future? There are four main design principles that are behind Industry 4.0 and support manufacturers and companies in implementing their technology-driven futures:
Interoperability: this is related to IoT and the transfer of data between communicating machines, devices, sensors and people. IoT is a concept with real-world integration already, and it will not be long until we see even fridges and everyday household items communicating with our handheld smart devices. IoT is an unavoidable necessity behind smart factories and Industry 4.0. Furthermore, IoT will continue the trend of automation, enabling automation to become smarter.
Information transparency: this is the ability of systems to generate virtual images of real-world manufacturing environments and then use real world data to change and improve this virtual image. Data will come from sensors and other cyber-physical devices. Raw data will need to be interpreted in order for it to become contextualised and relevant – this is no easy feat.
Technical assistance: this design feature has two parts. Firstly, the technology and information systems we deploy will need to be capable of informing humans and helping them to make important decisions related to the operations of the smart factory. This could be re-supplying or even making process change to enhance production and reduce waste. Secondly, assistance will be means of reducing human involvement in boring, repetitive, work unfit for humans, or unsafe work that can (and should) be automated.
Decentralised decision-making: AI will need to be able to make decentralised, independent decision, removing some of the brain work and managerial procedures from humans. This will mean humans will have more time to focus on important decision making.
Perhaps one of the most thought-provoking, alien, and fear-inducing aspects of Industry 4.0 is the last design feature, namely decentralised decision-making. The use of AI (artificial intelligence) has been subject to so much negativity in movies and films, it has polluted the true idea and application, and even the meaning of AI. Furthermore, there are those who seek to prevent job-losses due to AI systems.
That said, machine capability for decision making is so integral to Industry 4.0 it cannot be ignored. Factories are trying to achieve high performance in order to increase profit and reduce loss. Not only is there a profit gain to be made, but governments are increasingly tougher on manufacturers - honing in on industrial waste and energy consumption. We will see social and political factors reinforcing the need for efficient manufacturing. Humans have so far been largely unable to achieve this, in part due to our inability to make sense of data, or even collect that data until recently.
Today, our need to utilise and understand large quantities of data is inextricably tied into understanding current operating conditions and detecting faults, failures, and wastage. In production, there are tools that provide overall equipment effectiveness (OEE) information to factory management. These tools help to identify root causes. But in Industry 4.0, intelligent systems, such as IBM’s Watson (which is already active in manufacturing) will help to organise and interpret data. In fact, AI systems – built on neural networks – are advanced statistical machines, capable of adjusting weights depending on situational factors, or as it learns.
Industry 4.0 will use AI to assist in fault detection and identification, fault diagnosis, predictive analytics – whereby future issues can be identified and negated through data analysis. Factory maintenance can be, to some extent, automated. And where it cannot, AI will assist management in making important maintenance decisions or advance purchase of equipment based on predicted failure times. This will mean factories and producers will gain near zero level downtime.
The challenges of Industry 4.0
Although Industry 4.0 sounds like an almost inevitable future, there are numerous challenges to implementing its principles.
IT security issues are perhaps one of the most commonly cited reasons among management for not implementing Industry 4.0 principles. Opening a closed shop and filling it with communicating devices means that data flows are vulnerable to being hijacked and stolen. Even having data in the cloud raises huge data privacy issues for many. Related to security is the need to maintain company secrets in the manufacturing process. Opening data up and increasingly the likelihood of data loss can mean corporate espionage becomes more commonplace.
Reliability and stability are needed for critical machine-to-machine communication (M2M), including very short and stable latency times. Whilst installing sensors and cyber-physical devices sounds great, we also need to be smart about how these devices are deployed. It is not always evident how best to utilise these devices.
Upscaling IT infrastructure can be a considerable short-term cost, but it also poses long-term costs as critical devices will need to be maintained, and there will be employment and skills needs around critical IT capabilities. Skillsets are still lacking for Industry 4.0. Many countries are suffering with a lack of significant STEM (Science, Technology, Engineering, Mathematics) related skills and while this already impacts manufacturing, it will impact further when it comes to implementing and maintaining any IT systems for Industry 4.0.
Also important considerations are the general reluctance to change found within the manufacturing industry, with many old hats in the industry opposed to changes and digitalisation as it is seen as unnecessary, and the view that IT controlled processes are a threat to many. Job losses will naturally occur with the computerisation of manufacturing; there is a risk that certain skills will no longer be required. However, this can be mitigated through re-training workers so that their skills can be directed towards the maintenance of systems.
Implications for industrial gases
Processes in companies should be made more efficient, more effective and more flexible. Today, customers already expect that their special wishes are fulfilled with ever-increasing speed. This demands fundamental changes in production and logistics chains. In the smart factories of the future, machines, components and transport systems will communicate with one another. Classic intralogistics will and must evolve to a new level – smart intralogistics.
IoT can help to drive performance and relieve traditional supply and demand issues. IoT will be able to actively monitor the usage of industrial gases and, through the collection and analysis of big data in innovative AI systems being introduced to industry already, pinpoint areas of wastage and possible means of recovering that wastage through the application of new processes.
This means that, in the future, IoT technology may help to create supply chain processes that monitor the use of gases and automatically process an order to re-supply those gases when it becomes necessary given the most up-to-date usage statistics. This will not only help reduce costs, it will also ensure that companies don’t miss a beat when it comes to the particular processes that rely on important specialty gases. Whether in medical or manufacturing, all stand to gain time and money from IoT installations – the most valuable commodities in the modern world.
With the careful monitoring of supply needs, IoT will be able to amass huge amounts of data, which can later be interpreted by AI systems (as previously mentioned). But this has wider implications than just saving time and money – it will, in effect, enable humanity to better manage our limited supply of precious elements.