Volume 6 Issue 12 December 2024

A
Paperitalo
Publication
In this Issue

Welcome to Industree 4.0 for December 2024, exclusively sponsored by SAP.

SAP

By Nicole Smythe, SAP

How AI Insights Help Fantasy Football Fans Score Big

AI is revolutionizing fantasy football! Data analysis, predictive models, and optimization techniques empower managers to draft smarter, predict player performance, and optimize lineups, much like a supply chain manager. Discover how these same AI insights can boost your game and your business.


Does the phrase ‘waiting all day for Sunday night’ ring a bell? For those unaware, it’s the official theme song for NBC NFL Sunday Night Football, a musical beacon for millions each week as they tune in to watch their favorite team claw their way to victory. However, there’s another added factor to the NFL season – Fantasy Football. By this point in the season, millions of Fantasy Football ‘managers’ have drafted their stars and hope for a record-setting performance by one, or multiple, of their players.


And in today’s digital age, the modern-day technology often used behind-the-scenes to keep business running smoothly, are now available to amateur ‘managers’ everywhere. Through AI-powered processes, like data analysis, predictive models, and optimization techniques, fantasy football fans have the power to make better decisions and more positive outcomes, similar to a modern-day supply chain manager.


Let’s take a closer look at this intriguing overlap.


The Power of Insights: Merging Data Analytics with Game Strategy


There’s an old saying – ‘offense wins games, defense wins championships.’ Both are key foundational aspects to any sporting team, but it’s their harmonization that is key for success. The same goes for data analytics and artificial intelligence for both fantasy football management and supply chain management, alike. Often times, one can’t work without the other, rarely yielding success on their own accord. However, using both in tandem, people can analyze their way to an undefeated record and supply chain.


By leveraging advanced AI technologies, fantasy football managers, like me, can quickly analyze vast amounts of player performance data to make more informed decisions about their lineups. Within the official ESPN Fantasy Football mobile app, driven by smart AI technology, managers can see if their starter will be a ‘Boom’ or ‘Bust’ based on their projected points. This analysis comes from player trends to media buzz, giving managers smart insights and predictions.


Similarly, in supply chain, AI-driven data analytics can identify patterns and trends that help optimize inventory levels, reduce costs, and improve overall efficiency. Additionally, it can improve short- to long-term forecast accuracy through AI-powered algorithms, statistical modeling, demand sensing, giving supply chain managers the power to enhance their decision-making processes, which are crucial for agility and dynamically adjusting strategies.


Forecasting Success with Predictive Models


There’s a famous scene from Back to the Future II, when Marty McFly buys a sports almanac, which contains the score to every sporting event for the next, or in this case, last, 50 years. I doubt that there isn’t one football manager that doesn’t wish they could get their hands on that book, giving them the chance to predict the future. Well, in today’s modern age, we have something close – predictive models. These models are powerful tools that leverage statistical algorithms and machine learning techniques to forecast future events based on historical data.


In Fantasy Football, sites, like FantasyPros, use predictive models to analyze past player performance, injury history, and forecasted game conditions to project future performance, allowing managers to make data-driven decisions when setting their lineups, bargaining in potential trades, or deciding who to claim before the waivers. These models provide insights into player matchups, potential breakout candidates, and optimal draft strategies.


Learn more about how SAP Business AI can help your supply chain, and maybe even your fantasy football team, download our recent IDC InfoBrief: The Importance of AI in Supply Chain and Operations

Sustainable Automation

By Pat Dixon, PE, PMP


President of DPAS, (DPAS-INC.com)

While the 4th industrial era is defined by the ability to employ internet connectivity in industry, a parallel attribute of this era is the focus on sustainable operation. Today the term sustainable is not constrained to economic longevity. It also includes resource utilization and environmental impact.


For perspective, an article in IEEE Spectrum, December 2023, “The Transformer”, by Peter Fairley showed that under several projected plans there is a potential reduction of approximately 200 million tons per year of CO2 emissions in industry.  While transportation and power account for most of the CO2 reduction opportunity, industry presents a significant opportunity for reduced carbon emissions.


The paper industry is one of those industries. According to an article “Taking the Guesswork Out of Decarbonizing Pulp and Paper Mills” in Paper360º September/October 2024, the biggest opportunity for reduced energy consumption (which is connected with carbon emissions) is in paper drying. To move the needle will require research and development, along with significant capital investments in metal and concrete.  


Given that the digital domain of sensors, data analytics, and automation is the subject of this newsletter, what contribution could digital solutions in the paper industry make toward industrial CO2 emissions? You might be surprised at the answer.


Consider refining. Of course, outside of our industry refining means something you do with crude oil. In our case it is what you do to fibers to either get the shives out or develop them for bonding and runnability on a paper machine. Unfortunately, they tend to be black boxes. We know how much energy we put in, but we really don’t know what is happening in there. Lab samples once a shift doesn’t do much to detect a swing in fiber quality. Today there are online furnish quality systems that will tell you the quality of furnish and what the refiner does to it. This is especially pertinent to the growing utilization of recycled fiber. The impact of this measurement ability translates into automation. If these measurements can be used to stabilize operation by detecting and responding to swings, it can mean increased Overall Equipment Effectiveness (OEE) by reducing sheet breaks and running closer to optimal production rates. Any time you are running a mill without sellable product is lost tonnage and wasted energy. That wasted energy can be calculated with tools such as TAPPI TIP 0404-47. Calculations show that for North American production, measurement and automation for refining can theoretically reduce CO2 emission by 1.7 million tons per year.


That is not all we can do with online furnish data. If we assume the quality of a sheet of paper is deterministic, with enough pertinent data we can predict it. Often times a paper maker will add more basis weight in order to ensure the sheet meets those quality limits. Many of the quality limits are in terms of strength, such as tensile, tear, burst, and crush. An accurate dynamic prediction of these properties can be used in an optimizing multivariable predictive controller to choose the lowest cost and most sustainable target settings. This can make lowering basis weight possible by having other manipulated variables available. The reduced basis weight amounts to lower energy demands and CO2 reductions. Obviously, basis weight is also a quality target so there is a limit to low you can go. Another option is to develop grades at lower basis weight that maintain high strength. Online measurement systems that measure fibrillation can be critical to this capability. Fibrillation greatly increases the relative bonding area of fibers and therefore has a dramatic impact on strength. Very thin hairlike fibrils are called crill, and online measurement of crill is possible. Developing new grades at lower basis weights and high fibrillation can dramatically reduce energy demand. The combined result of these two lower basis weight solutions in North America can reduce CO2 demand by nearly 2 million tons per year.


Moving to the pulp side, causticizing efficiency and the lime kiln represents an opportunity for CO2 reduction. About 56% of North American pulp production uses chemical processes. Typical causticizing efficiency is 75%, which is below the theoretical limit of 82%. Online measurement of black liquor, white liquor, and green liquor in the recovery cycle pulping liquor makes it possible with automation to take standard 75% causticizing efficiency up near the theoretical limit of 82%. In practice, lime kilns in the industry are generating anywhere from 290.7 to 395.2 tons of CO2/day. Using a median figure of 342.95 tons CO2/day for a 1000 air dry ton/day pulp mill, improving causticizing efficiency to 82% would eliminate 106.95 tons CO2/day. The impact of this solution in North America is approximately 6.4 million tons of reduced CO2 per year.


Lastly, in the Kraft pulping process the objective is to remove sufficient lignin to separate individual wood fibers, which is measured with a test known as Kappa. Like with the lime kiln, there are not typically online measurements of Kappa number. Advanced controls in pulping use temperature and time calculations (H factor) and additional of chemical in proportion to production rate. However, variations in the nature of wood often mean additional chemical or heat are added to ensure Kappa does not get too high. Online measurement of Kappa allows controls to be tighter, which can reduce energy and chemical demand while boosting yield. The reduced chemical demand will provide additional savings to the downstream lime kiln, as described above. The Kraft process has an electricity demand of 2.34 MMBTU/air dry short ton and a steam demand of 19.3 MMBTU/air dry short ton (5) for a total of 22 MMBTU/air dry short. Using 59,433,206 air dried short tons pulp/year in North America kraft pulp production, if we increase yield by 1% due to a target shift of Kappa number, we reduce Kraft pulp demand by 594,332 air dried short tons pulp/year. This means a North American reduction of over 1 million tons CO2/year.


Adding all this up, there are over 11 million tons of annual reduced CO2 demand in North America through these digital solutions. There may be other pertinent digital solutions not accounted for in this article.  


We began by referring to an article stating that 200 million tons of CO2 reduction in North American industry is possible. It seems incredible that 11 million of that total could be accomplished in the paper industry alone without the high R&D and capital investment in metal and concrete. In the 4th industrial era, digital solutions can punch well above their weight class.

AI-We've only just begun

AI (Artificial Intelligence) has already been around a long time. For instance, “Spell Check” on your computer or phone is a form of artificial intelligence. However, a standard definition of AI, generally accepted by the public is elusive, even though ISO (the International Standards Organization) may have one.

 

As we move forward, our definitions of AI may become more contextual, needing an adjective to describe what we mean in certain applications.

 

Business AI may have many specific definitions. I am thinking in this way when it comes to just one little corner of Business AI (BAI).  A company may want BAI to draw on only internal intelligence, eschewing all that is known on a certain subject from all sources, the latter being a common way of thinking about AI.

 

Why would a business want their AI resources to be defined in such a limited way? Perhaps to stay out of legal trouble. Broad based AI may be able to go out and find exactly what competitors are charging for very similar products. While such knowledge is desirable, depending on its extent and what one does with it, it could be illegal in some countries.

 

And this opens another issue. Security—how does one protect their intellectual property from prying AI eyes? It may be resident in corners of the Internet where humans would not normally find it, but powerful AI can.

 

AI may be a bright shining light now, but we will need to proceed in a way it does not blind us. Rushing headlong into AI-land without careful consideration could be painful.

Technology set to 'extend' manufacturing capabilities

By Jack Lloyd

Described as an all-encompassing term for Augmented Reality (AR) and Virtual Reality (VR), XR technologies are becoming a source of value among many manufacturers. 

Read the full article here

Robots learn grasping from humans

By All About Industries

The IFL at the Karlsruhe Institute of Technology and the IAS at the University of Stuttgart are jointly developing a robot that learns human skills through imitation. For this purpose, they have established the ICM Future Lab Hapt-X-Deep, a research infrastructure unique in Germany.

Read the full article here

Enabling Real-Time Inventory Visibility with Advanced Systems

By Ellie Gabel

Advanced material handling systems have updated warehouse technologies by changing people’s workloads and giving them more time to devote to other parts of their roles, such as maintaining accurate inventory visibility. 

Read the full article here

The AI-Powered IoT Revolution: Are You Ready?

By Jay McCall

The convergence of AI and IoT promises to reshape industries and redefine the way we interact with the world around us.

Read the full article here
Industree 4.0 is exclusively sponsored by SAP