Adding intelligence to materials and products facilitates the fully decentralized operations model associated with Industry 4.0.
BY FRANCISCO ALMADA LOBO, Critical Manufacturing, Moreira da Maia, Portugal
Industry 4.0 is coming. It is the next major industrial revolution that will re-define manufacturing as we know it today. But what does Industry 4.0 bring to benefit an industry that already has highly advanced sophisticated manufacturing techniques?
The semiconductor industry is currently not one of those embracing Industry 4.0. Some of the reasons for this are based around the far reaching supply chain the industry uses, some because of the size of batches is still large in some businesses, and some because the idea of gathering greater quantities of information from machines is really not a new concept for the industry. To understand the benefits of Industry 4.0 to semiconductor production, let’s first look at exactly what it is.
A little about Industry 4.0
Industry 4.0 takes innovative developments that are available today and integrates them to produce a modern, smarter production model. It merges real and virtual worlds and is based on Cyber-physical Systems (CPS) and Cyber-physical Production Systems (CPPS), as show in FIGURE 1. The model was created to increase business agility, enable cost-effective production of customized products, lower overall production costs, enhance product quality and increase production efficiency. It brings with it new levels of automation and automated decision making that will mean faster responses to production needs and much greater efficiency.
FIGURE 1. Industry 4.0 merges real and virtual worlds and is based on Cyber-physical Systems (CPS) and Cyber-physical Production Systems (CPPS)
The Industry 4.0 model is inherently a de-centralized one with masses of data being transferred. The reduced cost of computer technology enables it to be embedded into shop floor materials and products. CPS then integrate computational networks with the surrounding physical world and its processes. Using the Industrial Internet of Things (IIoT), products will have the ability to collect and transmit data; communicate with equipment, and take intelligent routing decisions without the need for operator intervention. Cloud computing technology further gives a ready platform to store this data and make it freely available to systems surrounding it.
As CPPS compete to provide services to CPS devices a smart shop floor is created that acts as a marketplace. Adding communication and integration throughout the wider supply chain also means that different manufacturing facilities and even individual processes within a factory can compete for work; creating a Manufacturing as a Service (MaaS) model.
With hundreds of devices and shop floor entities producing information, Big Data and advanced analytics are also a major part of Industry 4.0. Simply collecting a lot of data doesn’t improve a factory’s performance. Advanced analytical software is needed to transform structured and unstructured data into intelligent, usable information. Having huge volumes of data also means this powerful software can be used to help predict production scenarios to further drive efficiencies and improve production strategy.
The intelligent operation and advanced analytics within Industry 4.0 will enable smarter decision making and provide the opportunity to further enhance processes. It will enable new products to be created, tested and introduced at a much faster rate with assured quality, consis- tency and reliability. The benefits are far reaching and so significant that this revolution will certainly come but the change will be gradual. To be sure not to be left behind, manufacturers will need to plan for the implementation of this predicted industrial revolution.
What does Industry 4.0 mean for semiconductor manufacturers?
For the semiconductor industry, the high cost of wafers make attaching electronic components to each wafer carrier or FOUP completely viable and presents huge benefits in increased production efficiency. Adding intelligence to materials and products facilitates the fully decentralized operations model associated with Industry 4.0 (FIGURE 2). With devices communicating with each other, the increased flexibility and productivity this model produces will make it possible to meet an increasing demand for greater manufacturing mixes and individualized products at much lower costs. For the production of semiconductors in particular, the very nature of the product being manufactured means there may also be opportunity and added benefit for some devices to hold their own information without the need for additional electronics. The information gathered from the decentralized model and analytical software used in Industry 4.0 also makes it easier to account for the cost of each item, resulting in better intelligence for business strategy and product pricing.
FIGURE 2. Adding intelligence to materials and products facilitates the fully decentralized operations model associated with Industry 4.0.
Although equipment used in the production of semiconductors already have sensors and transmit intelligent information into wider systems, the concept of the CPPS using the IoT adds a new level of simplicity to this idea. The cost of production within the semiconductor industry also means that even marginal variable improvements through the increased use of big data analytics will have huge financial benefits. The Internet of Things (IoT) will further enhance flexibility in measurement and actuation possibilities and free manufacturers from the time and cost associated with changes to sophisticated interfaces on production equipment.
The smart marketplace
With components interacting with machines and having the information they need within them about the processing steps they require, this creates a smart marketplace where the CPS requests services (demand) and the CPPS provides them (supply). Using mobile communications and cloud computing, this can of course be further expanded into the wider supply chain.
The concept of Manufacturing as a Service (MaaS) is, to some extent, already present in the semiconductor industry. The full supply chain has many different steps and, because of the high value of the product, transportation costs become pretty much irrelevant. This means that processing steps can be geographically distributed and the smart marketplace bidding for the work can extend throughout the world. Different factories may compete with each other for procuring specific processing steps and still be competitive regardless of location. Industry 4.0 gives the industry all the tools it needs for a smart, highly efficient marketplace that can add significant production flexibility while reducing both costs and production times.
Benefits of virtual and augmented reality
There are already few manual steps in the semiconductor production process with wafer production in particular using highly automated processes. This means there are few operators to oversee significant amounts of operations and equipment. Industry 4.0 opens up new areas in virtual reality (VR) and augmented reality (AR) that will help keep operations running smoothly.
The visualization and control of the wide spread autonomous elements within the CPS and CPPS in a decentralized production model requires a move away from standard, fixed, desk-top like workstations. Mobile devices are now more than capable of handling the demanding tasks of an operator workstation and offer the potential to decrease operational costs and increase productivity (FIGURES 3a and b).
FIGURE 3. Mobile devices are now more than capable of handling the demanding tasks of an operator workstation and offer the potential to decrease operational costs and increase productivity.
Using more comprehensive digital data and mobile computing technology, operators would be able to simply point a tablet at a piece of equipment and get real time information about what is happening. Locations of personnel could also be monitored to make most efficient use of human resources available. For the semiconductor industry; the use of secure, mobile devices further reduces the need to take up space in valuable clean-room environments.
Using mobile interfaces, maintenance technicians will also be able to conveniently move between machines without the need to logon at different workstations.
They can interact with different pieces of equipment and gather information about processes while carrying out tasks such as ordering spare parts all from a single mobile device. For specific operations relating to a piece of equipment, apps that automatically launch onto the technician’s tablet depending upon their location may further be used to add important additional infor- mation about a piece of equipment. For example, a particular part may be highlighted to be checked or replaced or additional information about specific machine readings highlighted on the display.
With all the amount of data sent by sensors, products and equipment it will also be possible to visualize in real-time the complete status of a production floor using VR 3D maps. Combining information about where personnel are within the factory and which direction they are facing, this further enables the implementation of some compelling AR scenarios. Indeed, the capability of mobile devices and the increase in real-time data available will likely make the wider use of both VR and AR a fundamental part of shop floor operations.
The Route to Industry 4.0 – the next generation of MES
There are a number of challenges that Industry 4.0 brings with it and its implementation will certainly not happen overnight. The huge benefits the model has to offer, however, can be planned into business strategies and realized over time. One of the first areas to consider is vertical integration of the model. This is important because corporate processes must not be avoided with the autonomy of materials and machines. Business processes for compliance, logistics, engineering, sales or operations all have components inside the plant as well as others that reside beyond the factory that are crucial to a business process being executed effectively. Without these, it’s almost impossible to properly manage a production floor of a certain complexity.
Modern Manufacturing Execution Systems (MES) based on decentralized logic offer a platform for the development of the Industry 4.0 model and a natural route to its vertical integration. MES have always been most effective when integrated into Enterprise Resource Planning (ERP) systems ‘above’ while monitoring and controlling production processes ‘below’.
With the CPS and CPPS communicating directly with each other, the MES can trigger business rules or workflows for the complete production process. For example, quality processes may demand that a device may need additional verification steps before processing continues as part of a higher level quality sampling strategy. This requires communication to intersect the business rules so the quality procedures are not bypassed before the device continues through its production processes.
Another area that is reliant on good vertical integration of systems within Industry 4.0 is Statistical Process Control (SPC). SPC requires data to be collected over time from numerous materials passing through the factory. For example, if a device within the CPS knows it needs to collect a measurable variable, this needs to be confirmed against SPC rules that it is within limits. If it is not, corrective action may be required. Flags for such actions need to be triggered in systems above the CPS and, again, the MES is an ideal platform for this.
By its very nature, the concept of a smart shop floor will generate huge volumes of data. An Industry 4.0 MES will need to aggregate this data and put it into a shop floor context. Indeed, to handle the decentralized logic and vertical integration of the autonomous entities on the shop floor, MES manufacturers need to fully expand their systems’ capabilities to ensure all plant activities are visible, coordinate, managed and accurately measured.
Future MES can also help to realize the full MaaS. This requires horizontal integration so all functions and services can be consumed by all entities on the shop floor including the CPS smart materials and CPPS smart machines. For individual equipment or processes to be procured in single steps, the MES needs to offer exceptional flexibility to expose all available services, capacity and future production plans. With visibility of the complete supply chain, MES also need to consider security and IP related challenges with multi-dimensional security. This needs to be at a service level but also at individual process, step and equipment levels and at any combination of these.
Ultimately it is envisioned that the Cloud will deliver the storage and the ‘anytime, anywhere’ ability to handle the volume of data created from sensors, processing and connectivity capability distributed throughout the plant. The manufacturing intelligence needed and provided by MES today therefore also has to expand to better accommodate the diversity and volume of big data. Fast response to any manufacturing issues will come from real-time analysis where advanced techniques such as “in-memory” and complex event processing may be used to drive operational efficiency even further, where the value of the process makes this a viable return on investment.
Support for advanced analytics in MES is needed to analyse historical data fully understand the performance of the manufacturing processes, quality of products and supply chain optimization. Analytics will also help by identifying inefficiencies based on historical data and pointing staff to corrective or preventive actions for those areas.
Legacy MES
Semiconductor was probably one of the first industries to embrace the idea of MES. First adopters were as early as the 1970s before the term ‘MES’ was even established. Some of these systems still exist today. The problem is that, as the limits of these early systems were reached; small applications have been added around them to meet modern manufacturing demands. These systems are so embedded into production processes that changing them is like replacing the heart of the factory and is no small consideration. There will, however, be some point where these systems can no longer be patched up to meet needs and factories will need to change to survive. The huge potential benefits Industry 4.0 offers may well be the catalyst to change and the basis of sound strategic planning for the future of a business.
Summary
One of the main areas of benefit of the Industry 4.0 decentralized model is the ability to individualize products efficiently with high quality results. This benefits all industries as trends show an increased demand for high mix, smaller batches to meet varying consumer demands. More than for many other industries, the high cost of individualized semiconductors makes the value of adding autonomy to customized processes even higher.
MES have been at the heart of the semiconductor industry for many decades but future-ready MES, based on models with de-centralized logic, offer a pathway to realizing the benefits Industry 4.0 has to offer. For semiconductors these benefits centre on reduced production costs, particularly for small production batches; enhanced efficiency of small workforces, and the business and cost reductions to be gained from the MaaS model and smart supply chain.
Although the semiconductor industry has been somewhat protected, competition is still fierce, especially in areas of mass production. In all different manufacturing areas, however, batch sizes will become smaller and the demand for individualized products will increase. Semiconductor manufacturers that can adapt more quickly to this trend will gain competitive edge and ultimately will be the businesses that survive and grow for the future. Without the Industry 4.0 model manufacturers will of course be able to produce in the future context of more customization, but costs will be much higher than for those who embrace this industrial revolution. If the full scope of Industry 4.0 is realized throughout the supply chain with MaaS, it will be even harder for companies that are outside of this model to compete in the smart marketplace.
With the dawn of Industry 4.0, manufacturing is moving into a new era that brings huge benefits and it is unlikely that the semiconductor industry will let itself be left behind!