Software with advanced machine learning algorithms, when combined with process design and operational know-how, have the potential to increase manufacturing efficiency and productivity in the process industries. Innovative business models can secure the sustainability of the achieved benefits by taking a more holistic view and focusing more on the goal as opposed to the means. It’s the outcome that matters and the evolving concept of Process Optimization as a Service offers a way to ensure long-term success.
The nature of manufacturing is changing across a wide range of sectors as businesses try to become leaner, more flexible, more proactive and asset light. In this eBook Hassan Yazdi, Product Manager for digital services at GEA, looks at how the manufacturing industries can use real-time process optimization as a service in this ultra-competitive world to maintain the highest levels of performance. Thanks to new digital service concepts, it’s now possible to challenge the established mindset of simply using Advanced Process Control (APC) solutions to augment existing systems and implement “Process Optimization as a Service” as a sustainable way to ensure continuous improvement.
APC for Process Optimization
Plant managers are continuously required to improve the productivity of their plants, applying advanced technologies to reach the highest possible production rates and improved availability with the lowest cost and environmental footprint. Propagating this demand down to the production departments has always meant investment in technology. Applying so-called APC solutions from software suppliers has traditionally been a common approach.
APC solutions are always challenged by the changing environments of production processes. The replacement of instruments with new measurement properties, degradation of equipment, new staff and the reconfiguration of production parameters and recipes are just a few of the natural but challenging deviations that APC systems experience.
APC user experiences across a variety of industries show the considerable benefits of these systems right after installation … but also note the slow degradation of the performance with time. A well-recognized global market report on APC solutions confirms that the “sustainability of APC solutions is [the] biggest user concern.”1 A major reason for this degradation is the lack of adaptation of the APC solutions to the changing environment. In response, some APC providers have tried so-called “adaptive” solutions; but, because of the variation of these changes in nature and magnitude, they have generally failed to close the gap. As a result, production managers see the APC solution to be running — but the outcome is often far from optimal.
A major reason for this degradation is the lack of adaptation of the APC solutions to the changing environment.
Why Not APC as a Service?
The concept of Software as a Service (SaaS) is well established and often defined as a software distribution model in which a third- party provider hosts an application and makes it available to customers via the Internet. SaaS is one of three main categories of cloud computing, alongside infrastructure as a service (IaaS) and platform as a service (PaaS). Such applications, also known as web- based software, on-demand software and hosted software, have become a common delivery model for many business processes.
It would be an easy segue in the manufacturing industry to apply this model to Advanced Process Control Software as a Service, which many suppliers will no doubt offer in the not-too-distant future; but, that is only part of the equation as this solution will only handle software-related maintenance aspects (Figure 1). Often, the reason for performance degradation resides in the process-related interactions between the product type, raw material, instrumentation and, to some extent, changing operator skills and experience. The latter seems to be an increasing challenge for plant management as it is difficult to retain skilled staff and ensure complete handover to new ones when they’re promoted to other jobs. So, if classic APC or SaaS-based APC is not the solution … then what is? In the following sections, it’s hypothesized that the ultimate outcome should be Process Optimization as a Service with a firm focus being placed on the “goal” rather than the “means.”
Learnings from Conventional Solutions
Common amongst almost all APC suppliers is a relatively high degree of technology expertise in advanced algorithm design … but a distinct lack of in-depth process and context know-how. Typically, these suppliers provide solutions to various industries, so their business model relies on providing general-purpose software that is commissioned for a specific plant with all the maintenance tasks being left to the in-house staff, who, most likely, have many other responsibilities and can’t monitor the performance on a full time basis and cannot do the necessary adjustment owing to a lack of the required skills.
Major producers with several facilities have already recognized this challenge and built up their own application development units; they typically buy the general-purpose APC software and develop their own solution and install it at the production sites — often based on an inter-company business model.
Regardless of the approach, the business model behind nearly all APC solutions so far is CAPEX based (Figure 2). You simply buy software and get it installed and commissioned at your plant. How to utilize the investment against the increasing demands placed on productivity and profitability is now plant management’s headache. The traditional CAPEX model relies on the expertise of a facility’s staff to get the best out of the system. Having a software update agreement on top of the CAPEX — or as a SaaS — improves performance in terms of the frequency of any updates.
"Software as a service" is only a part of the equation as it only handles software related maintenance aspects.
But, it still falls short of being able to reach the full potential for improvement potentials because the software updates cannot fully accommodate the full spectrum of changes that might be occurring within a plant; the agreement only ensures the availability of current software.
In contrast to the two business models, Process Optimization as a Service actually builds on the SaaS concept but also takes a holistic view of the infrastructure that the software exists in. As such, plant management benefits from being able to do much more than keep a certain program and its box up to date. Furthermore, such systems inherently reduce overheads and eliminate a certain amount of required maintenance as these aspects are all taken care of by the supplier.
It’s been done before, of course, but never in the process industries. How many companies buy software these days? The success of Microsoft 365, for example, allows companies to work with the most up-to-date applications for an annual fee. Updates are automatic. There is no “pride of ownership” of a computer program; people just need it to work.
Pharmaceutical businesses are increasingly choosing to streamline by using SaaS to enable their core business operations without the additional overhead of software installation, maintenance and support. Requiring no customer installation and no additional servers or other expensive hardware, the products can be made available quickly, providing pharmaceutical companies with unprecedented access to data and knowledge … without compromising the confidentiality of their own proprietary information. Dr James Willans, Chief Technology Officer at Lhasa Limited, has stated: “We believe SaaS is set to be a game changer for the pharmaceutical sector, delivering scalability and agility and lowering the cost of software that enables the delivery of safe, effective and life enhancing medicines and other products to market earlier.”
Likewise, PatSnap, a global intellectual property analytics company backed by Sequoia and used by innovative companies to accelerate R&D, now offers “chemical,” a SaaS platform that combines innovation data with vital and relevant scientific information into one single and easily searchable interface. chemical enables intellectual property professionals at science-led organizations to easily validate chemical development projects by analyzing big data through machine learning and artificial intelligence (AI). Its database links more than 114 million chemical structures, clinical trial information, regulatory details, toxicity data, more than 121 million patents and a number of other sources, quickly validating chemical development projects.
SaaS applications change the business and IT relationship completely, reducing the load on IT groups and enabling engineering departments to focus on mission-critical design and optimization activities such as engineering simulation areas.2
Rolls Royce has adopted an even more end-user value-centric framework with its jet engines, introducing the TotalCare® program two decades ago; customers are charged for flying hours and the system maintains a constant vigil over every engine to make sure it’s performing perfectly. In addition, many car companies are increasingly realizing that some people don’t actually want to own a car; they just need transport, so their business models have adapted accordingly. Grundfos, a market leader in pump manufacturing, is also going down the same route by changing its business model from being a pump supplier to a “pumping as a service” provider.
Many other examples exist; for example, no large company purchases printers any more. They don’t want to spend resources on maintenance and, instead, invest in “printing as a service.” Likewise, functions such as cleaning, maintenance and a wide range of office-based activities that had become a natural part of day-to-day business have now been outsourced to a service provider and relegated to the realm of Operations.
Common for all these examples is that the end user is now more focused on its main “goal” — managing the business — and can delegate the process to professional “service” providers. As previously stated, APC keeps us focused on the means, rather than the goals. It really doesn’t matter how advanced our APC solution is if we cannot meet or exceed our requirements for continuous optimization. The process industry needs to delegate this niche task to suppliers with the adequate context know-how and end-to- end responsivity to ensure the continuous high performance of its productivity.
Like many other suppliers, GEA has also supplied APC solutions to its customers using a CAPEX-based business model for many years. Focusing on the sustainability of these solutions, GEA has had to reassess its solution design and business model to incorporate the key components of industry 4.0 as well as process know-how- centered services.
The results of this journey were used to develop a new product — GEA OptiPartner — which was launched in 2019.3 The product is based on a SaaS concept in terms of its technology and business model but elevates the value proposition by utilizing process know-how to transform APC solutions into an end-to-end OPEX- focused Process Optimization as a Service product.
To ensure the continuous improvement of process performance, the product contains several elements that rely on process design knowledge and a thorough understanding of each customer’s operational activities. By using the technology to collect performance parameters and provide the expertise to analyze the data, the end-to-end supplier becomes actively involved in the production process, identifying areas for improvement, reducing unscheduled shutdowns and operating as an inclusive part of the process optimization team.
To ensure the continuous improvement of process performance, the product contains several elements that rely on process design knowledge and a thorough understanding of each customer’s operational activities.
International Dairy Magazine recently published an article titled “Intelligent Control Ensures Process Stability” that described how a DMK plant in Zeven, one of the largest milk dryers in Germany, was able to solve a series of serious problems using “a new, holistic and, above all, continuous process optimization” solution.4 More importantly, the article shows how continuously focusing on performance improvements can actually benefit the productivity of the plant in terms of production rate, quality consistency, energy saving and the give-away reduction of protein.
Focus on DMK Group story
- 87% reduction in process variations (measured as product quality variations)
- Closer on-target product quality and production rate increased up by 4.2%
- 4 performance review meetings with GEA recommendation on actions
- No un-planned shutdowns (even in summer 2018)
- On-line optimizers running at 98%uptime
Data Security and Intellectual Property
There is, of course, a natural and justified concern regarding data security in relation to Industry 4.0. Owing to demand for proactive services and the growing amount of data, cloud services are playing an important part in delivering next-generation services. Securing edge to cloud connectivity brings new challenges in the IT security space. However, the supplier is obliged to follow the highest standards of data governance, ensuring data ownership by customers and working in close co-operation with customers’ IT teams to maintain the highest level of data security.
In the GEA solution, data is transferred with a secure client- initiated connection following IoT best practice, meaning no ports must be opened into the network. The data is stored securely on Azure storage that leverages AES256 encryption. Data is segmented by the customer and access is granted on need basis using role-based access. To further increase our focus on data security, all access to the servers is monitored, logged and reviewed on a regular basis to decrease any risk of exposure.
Moreover, there must be a clear definition of what data is interfaced by the supplier to protect any data related to intellectual property (IP). Last but not least, the customer must receive comprehensive documentation on both the technical and data governance aspects to ensure an open and transparent insight into the responsibilities and confidence on both sides. The best results will derive from a combined effort with the customer’s IT team to ensure transparency about how the data are transferred and how to mitigate risks and eliminate security issues. With the focus now on privacy, data can be anonymized as much as possible and in full compliance with EU GDPR regulations.
Considering all of the above and owing to ever-changing market demands, there will always be a need to optimize plant. Up until now, the only solution has been a CAPEX-based investment in an APC model. Increasingly, the sustainability of this paradigm has been the user’s biggest concern. As a result, suppliers will need to modernize their business model to overcome this problem and provide an OPEX-based SaaS solution … but we’re not there yet. It is arguable that even this approach will not fulfill industry’s needs in terms of maximizing potential. So, the future success of these solutions is to build on SaaS and further develop an end-to-end OPEX-based solution, namely Process Optimization as a Service. In the end, it’s the outcome that matters.
With this business model, the relationship between the client and the supplier changes. In the past, the relationship was transactional, with an element of “after sales” service providing some level of continuity. With a service model, the supplier becomes more than just involved, it is completely committed to the ongoing success of the process.
With a service model, the supplier is completely committed to the ongoing success of the process.
- ARC Market Analysis, “Advanced Process Control and Online Optimization,” 3 August 2018: www.arcweb.com/market-studies/advanced-process-control-online-optimization.
- P. Reynolds, “Cloud Transforms Engineering into Agile Industrie4.0 Value Centers,” 2 August 2018 www.arcweb.com/blog/cloud-computing-transforms-engineering-agile-industrie-40-value-centers.
- GEA OptiPartner www.gea.com/en/products/gea-optipartner.jsp.
- “Intelligent Control Ensures Process Stability,” International Dairy Magazine 3–4, 20–22 (2019): www.blmedien.de/data/emags/moproweb/IDM_03_04_2019/page_20.html