The Intelligent Enterprise for Paper and Packaging Part 2: Using Industry 4.0 to organize production and sell more with smaller lot sizes
In Part 1 of the Intelligent Enterprise for Paper and Packaging series we talked about how digital technology is helping
paper and packaging companies redefine their core competencies to optimize their manufacturing operations, supply chains, and customer interaction to improve their products, services, and the customer experience. By 2025, a substantial part of paper and packaging companies' value, reputation, and differentiation will come from services. These services will be delivered around highly customized products, enriched by digital information. To get there, paper and packaging companies need to focus on five strategic priorities:
Offering small lot sizes and individualization
Running smart factories and digital networks
Supporting value-added services and new business models
Producing for purpose
Part 2 goes into detail on strategic priority 2: Offering small lot sizes and individualization.
Across all industries - and also in paper and packaging - Individuals and companies are interested in products and solutions that are built exactly to meet their requirements.
Often, these requirements are requested specifically to help differentiate them. The need to meet these requirements forces paper manufacturers to move from large lot sizes, which can serve many customer orders at low cost, to smaller lot sizes. As customers are not willing to pay more than for a standard solution, manufacturers need to control cost of manufacturing and distribution to stay competitive and profitable.
The Vision In 2025 companies will be able to quickly process small-lot-size sales orders and delivery of tailor-made solutions, creating additional value for customers. For example, instead of needing to order from predefined, inflexible solutions and delivery, customers will be able to request unique product features. Then, with help from rules, analytics, and algorithms, they will be guided to appropriate additional solution choices. Companies will offer much more flexibility on choice and delivery while still ensuring appropriate levels of profit. Being able to execute the wish of a customer from order through manufacturing and offering multiple delivery options - all while considering costs and margins - will be a key differentiator.
Manufacturers start toward this goal by getting all customer, product, production, and logistics-related data connected in a digital manner. They will use this data to analyze demand and improve production processes for higher flexibility and response times. Once all data is connected, digital technologies such as predictive analytics will recommend the best product configuration and the best way in which to fulfill its production and distribution.
Smurfit Kappa is a world-leading supplier of paper-based packaging solutions and is passionate about delivering quality products to its customers. That means producing the virgin and recycled papers that go into making its innovative and sustainable packaging solutions. With its unrivaled quality standards, Smurfit Kappa has harnessed data across its paper mills to drive consistent quality and productivity.
Providing solutions that precisely fit the needs of one single customer has been commonplace in traditional engineer-to-order environments. Now, manufacturers must be able to capture all customer requirements effectively and drive mass customization to give every customer exactly what they want.
Below are processes that can be dramatically improved with an Industry 4.0 approach. Companies can realize benefits including:
10-12% Reduction in total logistics costs
10-20% Increase in on-time deliveries
Up to 10% Reduction in total manufacturing costs
Improved Process Using Intelligent Technologies
Product variants predefined by marketing and product management
The customer requests unique product features and can be guided to appropriate choices
Variant product catalog with limited number of variants available
Product and production capabilities are managed with help from rules, analytics, and algorithms.
Every variant exists independently as a product, resulting in inconsistency and errors in variant management downstream
Machine learning technologies help to guide and identify the right configuration.
Unique design and engineering systems and bills of materials (BOMs) lead to high integration effort
Integrated design, engineering, and sourcing allows for quick order completion. Machine learning and image recognition help with ordering the right spare parts in case machinery fails.
Unique manufacturing BOMs and routing for each predefined variant, requires manual handling of BOMs, routings, and engineering changes
Production planning and execution are integrated for flexible and efficient operations. Manufacturing dashboards provide total insight, and management by exception helps keep focus on critical orders.
Inefficient, expensive, and error-prone variant manufacturing
High customer satisfaction results when customized products are delivered as quickly as standardized products
Critical for this transformation is the ability to manage the specifics of each order in every aspect of the industrial value chain in a consistent way at nearly the cost of a standard order. To do this, all product and process information must be kept in a single place, and all business processes - from initial engineering through after-sales service - must be effectively executed and closely monitored.
For more on how Industry 4.0 helps companies sell more with smaller lot sizes listen to our newest podcast.
In our next issue, we will talk about another strategic priority of paper and packaging companies.
Alfred Becker, SAP SE, Global Lead for Paper and Packaging
If you take all the data in the world and use it to train a neural network, the likely result is a neural network that learns noise. We live in a noisy world. Noise is not information and doesn't help you optimize your process.
It is no surprise that your process produces mountains of data. Machine Learning is a way to find gold in those hills of noise. Before explaining how, it is important to define our terms. What is Machine Learning?
The term Machine Learning is often used interchangeably with Artificial Intelligence. There are distinctions between the two. Artificial Intelligence is a broad term that refers to the ability of non-humans to perform as humans. Machine Learning is a subset of Artificial Intelligence using data to learn patterns and make predictions. The following illustration (found at https://discuss.analyticsvidhya.com/t/difference-between-ai-ml-and-ds/74724/2) helps explain these distinctions:
How can Machine Learning help you? I will offer an example.
Back in the 1980's there were attempts to measure the strength of a sheet of paper online at the reel. This would allow real-time measurement, and possibly control, of a critical quality attribute. This would eliminate the chance that an entire reel is produced, followed by samples sent to a lab, with later discovery that the entire reel you turned up, and the one you are currently making, does not meet specifications. Unfortunately, online measurement of strength properties continues to elude us.
Although we lack the direct measurement, we have a mountain of data that could predict the strength of that sheet to a reasonable degree. We know there are many measurements we have that affect strength. Machine Learning, properly applied, could provide a prediction of previously unmeasured parameters, which can make for a much more efficient and profitable operation.
That is one example of the potential of Machine Learning. To turn that potential into reality requires being able to discern truth from the noise of buzzwords and hype. Below are some tips:
While there may be some that claim they can take any and all data and turn it into gold, most of the work in Machine Learning is to take the noise out of the data to create a usable set of data for Machine Learning to train with. This is called pre-processing. Applications that lack the ability to help you pre-process data should be met with skepticism.
Aligning data in time is part of the pre-processing. The moisture that is measured at the reel affects strength immediately, but the basis weight valve is many minutes before the reel, and the refiners are further back. Knowing what affects the prediction you are trying to make, and when, it imperative to create a usable model.
If you are only trying to create an accurate prediction, you can be a little less picky about the dataset you create. If you are using Machine Learning to help you optimize or automatically control your process, you need to narrow down your dataset further. The inputs in your model should be independent of each other and have significant effect on the measurement. Failure to remove redundant or insignificant inputs can result in nonsensical predictions and behavior.
A big advantage of a Machine Learning algorithm (such as a neural network) is that it can model a relationship of nearly any complexity. That is also a danger. A model used for optimization or control that has unreasonable relationships has essentially learned noise. While we have plenty on nonlinear steady state relationships in our processes, we know that most of them will have fewer than 3 inflection points. Although neural networks are a black box technique, it is possible to show gains (sensitivities) to reveal whether the model is realistic. This is a feature that I regard as essential in a Machine Learning product for an optimization or control application.
Machine Learning efforts can benefit from a data scientist that knows how the algorithm works and what helps to create an accurate model. However, this alone would likely spell disaster. Without process knowledge, your prediction might be useless. Knowing how a process behaves and applying first principles cannot be replaced with data and algorithms. Combining a process expert with a data scientist can be an ideal team to produce successful Machine Learning results.
Just like all of use, Machine Learning predictions will not be perfect. There will be mistakes. Setting expectations and knowing how to handle situations where the prediction leads the wrong way needs to be considered to maintain confidence in the application.
To mitigate erroneous predictions, use your lab. It has not been rendered obsolete. You cannot measure everything. A furnish change (hardwood, softwood, recycle) may affect the strength of your sheet without a clear measurement in your system to account for the composition of fibers in your sheet. Using your lab as feedback to bias your prediction is a good practice.
This is a sampling of tips. There are many more, and experience is a great teacher. Creating your own datasets and playing with available products or some of the freeware that is available (such as Python and R) can help you get more familiar with the triumphs and pitfalls.
Machine Learning is being applied in our industry and producing gold. It can work. It can also fail if hype overshadows reality. Filtering the noise from the hype can help mitigate the noise in your Machine Learning investment.
Pat Dixon is Southwest Region Engineering Manager for Global Process Automation (GPA), a controls system integration firm.
Is My Smart Factory Smarter Than Yours? It's Hard to Say
The smart factory. Industry 4.0. Smart manufacturing.
What those terms mean is something of an open question. Each has become more of a sort of marketing slogan with evolving meanings than a definable signpost for high-tech manufacturing. And terms referencing intelligence or, for that matter, industrial revolutions, tend to become ever-more grandiose or more generic over time.
Such is the case with the second of the terms in this list, which came to life in 2011 as a German-government-backed Industrie 4.0 initiative to help the nation's manufacturing sector maintain its competitive edge. The phrase itself hints at a cyberphysical, systems-driven industrial revolution, but, with its ".0" suffix, points to the level of change one would find in an incremental, but substantial software upgrade. It also conjures to mind the Web 2.0 that largely receded from memory since it popped up around the turn of the century and peaked around 2007. And at this rate, Web 3.0 has failed to become a mainstream buzzword.
As the adoption of Internet of Things (
IoT) devices continues to rise, new research from Irdeto has revealed that eight in ten organizations have experienced a cyberattack on their IoT devices within the past 12 months.
Of the organizations whose IoT devices fell victim to a cyberattack, 90 percent experienced an impact as a result including operational downtime and compromised customer data or end-user safety.
Why Industry 4.0's potential needs to be proven on the shop floor
99% of mid-market manufacturing executives are familiar with Industry 4.0, yet only 5% are currently implementing or have implemented an Industry 4.0 strategy
Investing in upgrading existing machinery, replacing fully depreciated machines with next-generation smart, connected production equipment, and adopting real-time monitoring including Manufacturing Execution Systems (MES) are manufacturers' top three priorities based on interviews with them
Mid-market manufacturers getting the most value out of Industry 4.0 excel at orchestrating a variety of technologies to find new ways to excel at product quality, improve shop floor productivity, meet delivery dates, and control costs
Real-time monitoring is gaining momentum to improve order cycle times, troubleshoot quality problems, improve schedule accuracy, and support track-and-trace
The survey was conducted by Market Measurement, Inc., an independent market research consulting firm. The survey included 230 executives at U.S. manufacturing companies with annual revenues between $200M and $3B and was conducted in November and December of 2018. Please see page 2 of the study for additional details regarding the methodology. One of the most valuable findings of the study is that mid-market manufacturers need more evidence of Industry 4.0, delivering improved supply chain performance, quality, and shop floor productivity.
AI in Business: Creating intelligent spaces with IoT
Tucked away in a corner of a quiet neighborhood in West Hartford, CT, French electronics manufacturer Legrand operates a massive factory that churns out an array of lighting, wiring, and building supplies.
"You'll fine Legrand products in the electrical infrastructure within a building," explained Brian DiBella, president of Electrical Wiring Systems at Legrand. "There are primarily three units that work on that here in North America. We have wiring devices, which includes switches, outlets, dimmers. We have a cable management division that provides access to power or data, either in or on a wall or in or on the floor. And then we have an overhead systems business that makes large-scale cable trays that you'd see in commercial and industrial buildings.
Founded in 1865 in Limoges, France, Legrand began to focus on Internet of Things integrations in the last five years. "We deliver power, light and data to millions of spaces all over the world," said Manny Linhares, Jr., director of IoT Strategy for Legrand North and Central America. "We really try to focus on the user experience and improving that, and over the last five or so years, we've really put a lot of time, effort, and energy into IoT."
One of the strengths of internet of things (IoT) technology is that it can do so many things well. From smart toothbrushes to predictive maintenance on jetliners, the IoT has more use cases than you can count. The result is that various IoT uses cases require optimization for particular characteristics, from cost to speed to long life, as well as myriad others.
But in a recent post, "
How the internet of things will change advertising" (which you should definitely read), the always-insightful Stacy Higginbotham tossed in a line that I can't stop thinking about: "It's crucial that the IoT optimizes for trust."