Top 5 AI Predictions for Manufacturing in 2022 (And How to Prepare Today)

The year that has gone by—2021 was an eye-opener at many levels. Both process industries and discrete manufacturers have woken up to the crucial benefits of AI implementation. It is now a proven fact that leveraging Big Data and AI can facilitate the detection of patterns, accurate interpretations, and real-time communication of actionable insights to catalyze efficient workflows, faster decision-making, and optimized production.  

In 2022, Big Data and AI apps will undoubtedly impact how manufacturers do business, but what impressive strides will the powerful technologies take in the upcoming New Year? 

Top 5 AI Predictions for Manufacturing from experts: 

  1. Monetization of manufacturing data will be at the crux of agile manufacturing. 

Organizations that maximize their investments in data now will be the early birds to catch the business benefits and rule the roost ahead of their competitors. Stronger data teams and AI-enabled integrated data analytics will streamline processes for optimal efficiency in operations.   

Aside from that, data-drivennetworking between organizations will become the norm as AI apps will enable safe interoperability and data exchange. Ethical acquisition and exploitation of data, not only within but also across organizations, will establish data alliances like never before.  

The monetization of manufacturing data is the process that provides a measurable economic advantage by using (aggregated and transformed) manufacturing data to exploit data-driven knowledge spillover effects. AI apps can help shed inhibitions around data privacy. Strategies for monetizing manufacturing data can be manifold, offering an added economic value of more than $100 billion is predicted. [1] 

  1. A more holistic approach by factories toward AI adoption will prove to be a veritable game-changer. 

Factories dipping their toes and testing the waters will take the plunge. Instead of working in phases with incorporating AI applications in their processes, the push will be toward a system-wide change. Realized intelligent factories using integrated data from all the assets—operational and human—in the manufacturing network will become prevalent features. The gap between AI leaders and the companies following in step will get mitigated with accelerated digitalization. In fact, we can expect organizations to work with a blueprint, clearly prioritizing AI initiatives for business value and definitive manufacturing goals. 

  1. Green AI adoption for sustainability initiatives will be the mark of a responsible manufacturer. 

With the help of AI, companies will make a conscious effort of switching to environmentally sustainable manufacturing. The shift from a manufacturer-centric approach to eco-centric decision making will turn the industry on its head, allowing more responsible players to get into the front row.  

AI applications can offer tremendous new possibilities to help manufacturers adopt green manufacturing methods and reduce their carbon footprint. AI can optimize and ensure responsible energy consumption in manufacturing facilities and reduce wastage of natural resources. Integrating data analytics and AI in the manufacturing processes can facilitate the designing of products for sustainability by taking into account the environmental impact at every step of the product lifecycle.  

  1. Reinforcement learning will bring a radical transformation to the manufacturing scenario. 

Reinforcement learning (RL), an unsupervised machine learning algorithm, will deliver a transformative advancement in 2022.  

AI applications will get more intelligent with the simulation-based dynamic programming method. RL algorithms converge toward optimal process control strategies with data-driven solutions to problems in the production domains, enabling an autonomous manufacturing system. The key will be having suitable data sets to train algorithms, which requires a culture of data sharing within companies. Then machine learning can step out of the lab, where the training is slow and expensive, and get into production environments, where things get real. 

As it trains on the shop floor, instead of working with pre-programmed training modules, the complex real-time evaluation helps the AI applications to determine which actions are suitable in the long term. Reinforcement learning can thus bring a turning point in the world of manufacturing.  

Read about RL in detail HERE

  1. The Internet of Behavior (IoB) will enhance customer-centricity 

AI applications can help organizations collect and analyze behavioral data of employees engaged on the shop floor. The collection and usage of that data to influence future behavior is called the Internet of Behavior (IoB). As organizations improve the quality of data captured, they can also combine data from various sources and leverage IoB to improve the process workflow or influence interaction with customers, understanding their needs, overcoming challenges, and deriving customer-specific solutions. 

As per Gartner, IoB will tightly link customer experience and employee experience to transform the business outcome. It can help differentiate a business from competitors, creating a sustainable advantage. This trend enables organizations to capitalize on COVID-19 disruptors, including remote work, mobile, virtual, and distributed customers. [2] 

That is how we see things shaping up!  

As the world transitions from the pandemic to a state of normality, 36 % of manufacturers say they are currently engaged in AI projects, and 23 % more are planning to use AI in the coming months to unlock $13 trillion in value that industry experts anticipate from industrial sectors.  Manufacturers learned a lot during the pandemic and are now looking for the best way to become more productive, safer, and more agile by leveraging the petabytes of data insights harvested from connected factories. [3] 

We hope this article gives you a great insight into where the industry is heading in 2022. 

References: 

[1] Trauth, D. (2020, July 27). Monetization of manufacturing data. Senseering. https://medium.com/senseering/monetization-of-manufacturing-data-7e55d4c213ed 

[2] Panetta, K. (2020, October 19). Gartner Top Strategic Technology Trends for 2021. Www.gartner.com. https://www.gartner.com/smarterwithgartner/gartner-top-strategic-technology-trends-for-2021 

[3] IOT World Today. (2021, November 9). AI led Digital Transformation of Manufacturing: Time is NOW. IoT World Today. https://www.iotworldtoday.com/webinar/ai-led-digital-transformation-of-manufacturing-time-is-now/ 

Reinforcement Learning, the Effective Solution for Process Control Optimization.

Gradual and deliberate ‘change for the better’ eventually improves productivity and facilitates efficient processes.

Kaizen, the Sino-Japanese business philosophy, urges you to strive for continuous improvement. Applying the principles of Kaizen to manufacturing focuses on improving the four main elements of manufacturing — product, process, people, and environment. Kiichiro Toyoda, the founder of Toyota Motor Corporation, championed the philosophy by saying, “The ideal condition for making things gets created when machines, facilities, and people work together to add value without generating any waste.” [1] 

Harnessing the power of data analytics thus becomes crucial to a company’s continual process improvement strategy. Manufacturing process control optimization drives you toward achieving several business goals, from improved production efficiency to reduced production costs and increased customer satisfaction. However, integrated data analytics by solely relying on limited human capabilities is not only a terrible challenge but can also be misleading and flawed. Measurement and monitoring of Key Performance Indicators (KPIs) are essential for process control, but with limited analysis capabilities, it gets difficult to control wastage completely.  

The manufacturing sector has thus undergone a radical transformation over the last few years with the advent of Artificial Intelligence (AI). Machine learning is a subfield of AI that enables machines to become intelligent with precise data analytics. It makes manufacturing process optimization possible by utilizing a simulation-based dynamic programming method called Reinforcement Learning (RL). 

Reinforcement learning is an unsupervised machine learning algorithm that has revolutionized the continuous manufacturing scenario. It enables an autonomous system that learns from interactions with the industrial environment in real-time, trial and error, and does not rely on preprogrammed training. RL catalyzes SMART manufacturing by learning from accomplished tasks, incorporating changes in behavior based on feedback about the results of previous actions.  

The biggest challenge for manufacturing optimization lies in collecting and cleaning vast amounts of data from different sensors every day while maintaining consistent uptime of production facilities. In many manufacturing systems, continuous state variables and discrete actions get detected. Therefore, process optimization becomes a highly complex task. Breaking down siloed data is essential and controllable parameters need monitoring and adjustment as they directly affect the production throughput.  

Though these nonlinear dynamics can make process optimization challenging, it is not that difficult for RL algorithms to converge toward optimal control policies. Since experimentation on the shop floor could get expensive, the best action path gets derived through simulation. A digital twin of the physical production system gets created by domain experts who understand how the system works. The algorithm then learns the complex relations between the various parameters besides their effect on production and output.  The feedback loop helps the goal-oriented algorithm learn and make sequential decisions while building a framework for the ideal process flow to generate desired results. In operations, RL-based approaches have delivered consistently impressive results.  

Manufacturing is a complex system, where failure at balancing the customer’s demands with production costs, inventory, batch-cycle time, and quality can result in catastrophic outcomes. Manufacturers need to take a data-driven approach while developing robust, adaptive, and scalable solutions. Reinforcement Learning is an effective tool for predicting and solving specific problems in manufacturing processes. In fact, today 40% of all potential value created by data analytics comes from AI and ML techniques. In totality, machine learning can account for about $3.5 trillion to $5.8 trillion in the annual value — as per a McKinsey report. [2] 

To summarize, RL enables data-driven and flexible processes in production domains. It has, therefore, already been deployed in production systems, proving to be an effective solution for process control optimization. 

Myths about Digital Transformation/Manufacturing.

What is a myth, but a traditional narrative that catches people’s imaginations and soothes their sentiments! The stories get told with much conviction, with minimal or no corroborative proof, and soon become a widespread belief. Business myths are misconceptions arising from the assumptions made by ill-informed minds. Mythological tales can be very entertaining and form a part of a rich culture. Business myths might not serve any purpose and can be misleading, discouraging, and detrimental to company culture, too. If only we were to read between the lines, we would discover truths. 

Ever since tales of the Industry 4.0 became rife, innumerable ideas — true and false — regarding the digital transformation of manufacturing have started doing the rounds.  

Here are some myths we thought needed busting so that you can make informed decisions based on facts: 

Myth #1: Digital transformation is a big-budget affair, therefore, not meant for small companies.  

Truth:  

The first thought that comes to mind when discussing digital transformation is the costs involved. As nothing comes for free, digitalization comes at a cost, but strategically planning an AI adoption strategy will not drill a hole in your pocket. On the contrary, the resultant business benefits in terms of ROI will outweigh all expenses. Phasing out the transition could also help SMEs streamline their processes and enjoy profits. Industrialists have been reinvesting their gains into leveling up on the digital platform. AI applications for manufacturing optimization are accessible and affordable by all.  

Myth #2: AI is a disruptive technology that will remove the ‘man’ from the equation in manufacturing.  

Truth:  

All of the previous industrial revolutions since the late 17th century changed our life drastically.  History stands as proof that technological development has only created more jobs and better opportunities for humans to progress. AI is yet another technology designed to enable human capabilities and not cripple society. With AI assistance, humans can increase productivity by completing job tasks faster with precision accuracy, making businesses profitable. That will also leave more scope for skilled workers to get employed in more meaningful roles.  

Myth #3: The outcome of Smart Manufacturing is not measurable in tangible values. The benefits are not immediate.  

Truth: 

 For an average USD 5 billion company with a 10% margin, investments in digital technologies produce an additional USD 425 million in profit. Of all surveyed industrialists, more than 60% have stated that digital transformation can help organizations address the business objectives of prime importance — reducing operational costs and growing market share organically. [1] 

Neewee’s own ready-to-deploy AI apps can be integrated with any IoT platform to streamline the manufacturing process end to end, deliver higher ROI and improve product quality within six to eight weeks of implementation. We do this by creating Process Digital Twins that connect and mirror the manufacturing life cycle, revealing cause-and-effect relationships between components, raw materials, and processes. The apps then provide predictions and actionable recommendations using Machine Learning and AI.  

  • Improving process yield by 8-10% 
  • Enhanced product quality by 10-12% 
  • Reduced working capital by 8-10% 
  • Cut maintenance costs by 15-20% 

Myth #4: You can take your own sweet time to digitalize your manufacturing processes. Also, it’s optional!  

Truth:  

The general tendency is to wait and watch how the early adopters of technology are faring with the new intelligent manufacturing methods. However, sitting on the fence is not a healthy strategy if you wish to stay relevant in the rapidly increasing competition. A digitally connected lean manufacturing eliminates wastage, reduces batch-cycle times, and repeatedly delivers consistent quality and throughput. This directly translates as increased customer satisfaction. While traditional ‘unintelligent’ linear manufacturing systems will continue to struggle with delivering on promises, which can end up putting off the customers. Modernization of manufacturing is no longer an option but an imperative for business growth. 

Harvard Business Review sums up: “By the time a late adopter has done all the necessary preparation, earlier adopters will have taken considerable market share; they’ll be able to operate at substantially lower costs with better performance. In short, the winners may take all and late adopters may never catch up.” [2] 

Myth #5: For successful AI adoption in manufacturing, you need to digitalize the entire factory in one go! 

Truth:  

Big Data Analytics, Industrial Internet of Things, Machine Learning (ML), and Artificial Intelligence (AI) apps are different aspects that work together to turn your business operations toward Smart Manufacturing. Each involves cost and complexities. It is neither advisable nor practical to digitize the entire facility in one go. AI applications implementation into the system, working with integrated data analytics, is a step-by-step process involving reiterations to arrive at the best fit for optimizing different production processes. Since there can never be a one-formula-fits-all solution, only a well-developed strategy for AI Adoption guarantees astonishing results.   

It is not debatable anymore, manufacturers can optimize their business to achieve agility, efficiency, quality, and sustainability with AI Applications.  

Final Takeaway: 

The hype cycle about digital transformation of manufacturing carried many myths and even more facts. We have debunked some myths that might be keeping many from taking timely decisions for a digital strategy. It is therefore essential to confirm the credibility of the source before you believe the bit of news. Taking decisions based on myths would be like that notorious game of Chinese Whispers. You run the risk of falling for inaccurately transmitted gossip. 

Who is the SMART and Savvy Manufacturer?

S.M.A.R.T. The Oxford dictionary offers some definitions of what the adjective means. We all have grown up grooming our self to be just that. However, as the decades flew by, the meaning of the word kept evolving. 

In 1981, three management gurus, George Doran, Arthur Miller, and James Cunningham, published an article that changed the simple word into an acronym. They said that our goals had to be Specific, Measurable, Assignable, Realistic, and Time-related. “Sure,” we all said. Then the wheels of time turned some more and taught organizations that the letter ‘A’ had another role — Automation. And now, we are in the era of Industrial 4.0 where SMART has become the buzzword once again. This time it has a whole new definition, especially in the context of manufacturing.  

The new SMART is all about Augmentation. It coaxes you to adopt Augmented Intelligence to get to your manufacturing goals. Specifically, it is about using Artificial Intelligence (AI) technology — machine learning and integrated data analytics — not as a disruptive force but to enhance human intelligence.  

Customers today demand and settle for only the best, fitting their exact specifications, and they want everything delivered as scheduled. Manufacturers are feeling the pressure to accelerate and optimize production to stand up to customer expectations. Those at the forefront of intelligent manufacturing are redefining their production processes and becoming more efficient with digitized, networked production sites. Their goals are specific and measurable ROIs, realistic and time-sensitive customer satisfaction. 

Consider this recent use case scenario. Recently one of India’s largest defense aviation companies, which also works in the civil aviation space, was scheduled to supply critical component assemblies to a global aviation OEM. The complicated production process of the components required a specific raw material — Titanium plating. Besides, some parts also had to be exported for process completion. One can easily imagine the pressure! The monthly production volumes were high, and delivery timing was strict and contract-bound.  

To state the business problem: Their production plan got created at the Macro level once every quarter and reviewed in detail only once a month, leaving no scope to factor in the delays that the raw material suppliers could create or the variability of lead times for the special processing of the components in foreign countries. Their detailed production plan also sorely lacked clarity of daily production issues such as resource availability, shift adjustments, or even jig breakdowns. The cumulative impact of these process anomalies resulted in delayed production and failure in adherence to delivery commitments stipulated in the contract. Not to mention the significant losses incurred due to the increased WIP inventory, poor delivery rate, penalties for non-compliance with contract terms, and decreased customer satisfaction.  

Fortunately, the manufacturer took a timely leap, switching from traditional automation methods to an AI-enabled transparent, flexible, and efficient manufacturing ecosystem. And where there is theBodhee®Integrated Micro Planning & WIP Reduction Appthere are real-time actionable solutions and direct business benefits.  

The silos got broken down, and the cleaned-up master data inputs enabled integrated analytics. The app identified the gaps in raw material supply chain management. The algorithms calculated the actual lead times for process completion and recalibrated the production system. Constraints, as well as production goals, were configured for multi-objective scheduling. A detailed plan got customized considering every possible variability, from suppliers to lead times for the parts that needed special processing and resource availability, production capacity, factory holidays, staff shifts, and other shop-floor realities such as repetitive quality issues. The detailed production plans were made accessible to the production team with real-time actionable insights for timely adjustments. That ensured a significant reduction in WIP and efficient workflows, which translated as delivery targets getting achieved.  

Within just 8-12 weeks, the Tier-1 aero-component supplier began enjoying reduced working capital by 8-10%, 5-7% improved capacity utilization, and 3-5% reduced batch cycle time.  

So, now we know the answer to, “Who is the SMART manufacturer?”  

The one who leverages data through integrated analytics to generate golden ROI is a SMART manufacturer. Also, the one whose goal is to break through the competitive manufacturing landscape and emerge as a new market leader. Savvy?   

How IIOT Platforms are Getting Impacted by the Idea of Use Cases Driving Scalability and Accelerating ROI.

AI and use case-based applications are the future of IIoT, say experts and analystsWebinar Report

In the world of multi-dimensional data today, the traditional asset-centric and industry analytics IoT platforms can no longer provide competitive benefits or business impact. Thus, organizations are increasingly adopting Industry 4.0 and digitalizing their entire manufacturing value chain. Yet, while making this shift, many cannot understand the value and importance of the immediate impact that Digital Industrial Platforms can have on their manufacturing. 

In a virtual conference on ‘The Emergence of Use Case-Based Applications on Digital Industrial Platforms’ conducted on Tuesday 7th September 2021, industry experts and thought leaders joined us as our esteemed speakers. Dr. Marie-Isabelle Penet — Global Industrial Operation Excellence & Transformation Manager at Euro API (Sanofi) along with Dr. Zoltan Finta, Digitalization Global Leader at Euro API (Sanofi), Dr. Paul Miller — Principal Analyst at Forrester, and Mr. Jaspreet Bindra, Thought Leader in AI and Digital Transformation together with Neewee’s very own Co-Founder & CEO Mr. Harsimrat Bhasin discussed the future of Digital Industrial Platforms and their importance, and the generation of quicker ROI by the use case-based applications approach. 

Mr. Harsimrat Bhasin opened the forum to invite featured guest, Dr. Paul Miller. Based in the UK, his area of research coverage is the Smart manufacturing space. Miller astutely directed attention toward a 2015 McKinsey report that said 80 – 90% of industrial IOT pilots failed to scale. He also observed that the organizational issues and processes around it were problematic — not the technology. Thankfully, there was a paradigm shift. The formerly mum manufacturers began speaking up about their business problems in improving their processes for better yield rates, understanding how to deploy predictive maintenance, and focused on solving problems.  

Dr. Marie-Isabelle Penet and Dr. Zoltan Finta shared their real-world experience of overcoming their manufacturing problems and optimizing production with the help of Neewee. Penet spoke about how Euro API, being a part of a big company like Sanofi, made it imperative to manufacture consistent quality Active Pharmaceutical Ingredients (API) on a big scale for Europe. The challenge was to identify correct process parameters and control them while also improving productivity and quality profiles.  

When asked why they turned to Neewee in particular, Finta said, “We wanted to work with a company that has extensive experience in manufacturing digitalization. Because Neewee is flexible enough to not only understand chemistry but also understand our most specific problems; dedicated to working together on those problems relentlessly until we find a solution.” He also added, “With Neewee, it is not a connection as a customer and the supplier. No, this is a partnership.” 

Penet confirmed that Neewee’s structured approach afforded clarity on what to do, what results to expect; they saw measurable benefits within three months. 

“Major benefits of the collaboration with Neewee is that there are actionable insights for implementation, and they always conduct work with maximum transparency. We have had not only one collaboration, but we did it again,” she said with a broad smile. 

However, Finta said it was challenging to convince the management at their manufacturing plant. The experts with 20-30 years of industry experience came to the table believing that they knew everything about API manufacturing and that there was no scope or requirement for digital transformation. 

“Thanks to the collaboration culture with Neewee, there was a total openness in discussions of any idea or feedback. Their agile response to anything the manufacturing site needed. We only needed to wait for the site representative to recalculate and realize they had reached maximized results, quantified as maximum yield. The management then committed to digitalization.” 

Bhasin explained the Neewee perspective, “It’s having a shared goal, which is very important. What will be the ROI at the end of it? Because then you know what you are working toward. We are not experts in their chemical production processes. So, what we bring is the way of looking at data, which can drive that value. Collaboration helps to scale it up and make it a sustainable long-term success. When you quickly identify problems in production processes and show them hidden values, it’s the proverbial tasting blood. It becomes self-perpetuating at that point.”  

Miller made it a thought-provoking discussion by asking Mr. Jaspreet Bindra to comment on failing fast at scaling success.  

“Successful digital transformations are successful experiments. Successful for many reasons, but failing for only one reason — lack of the organizational culture mindset.” Said Bindra. “A company becomes digital when it starts thinking and behaving in a different way. Business models must change to focus not only on the technology but also on customer problems,” Bindra added. 

As the webinar drew to a close, Bhasin spoke about how Neewee has helped Sanofi with initial success and digital transformation at a few plants and what is being done to replicate it globally across the entire organization in the future. 

“One is to have a solution that can work seamlessly across different sites similarly. Second, giving the same level of confidence to the management that we can deploy solutions rapidly. And last, giving on-ground actual users actionable insights,” Bhasin replied perceptively, adding, “Paul, sometimes you have to start small to show the success. Confidence in newer ideas builds upon success, and then it grows in circles, getting bigger and bigger!” 

The webinar ended on a high note with Finta’s words, who spoke like a true visionary, “Machine learning and artificial intelligence is a part of our digital transformation. I think that the future is to scale up what we started together with Neewee. The key element is to get reliable master data for data analysis. And instead of just one or two projects, we need to think 20 or 25. When the different digital tools get connected, we could use the same collected data for different purposes. We must not only scale up the fund to expand the portfolio but also increase it, to finally embrace the digital industrial platform in its entirety!”