An Eye on
An Eye on
An Eye on Smart Manufacturing
Why do innovative ideas get stuck in pilot purgatory, and how can manufacturers set themselves of up for IoT success?
In today’s connected world, manufacturing remains far from untouched as waves of change from IoT (Internet of Things) technologies crash over society and industry. Smart manufacturing refers to the digitization and connectedness of manufacturing, creating a fully integrated network of all aspects of production—including fabrication, supply chain, logistics, and beyond. Smart manufacturing also means making use of information generated from data gathered from manufacturing processes to make decisions in a systematic and holistic manner that improves speed, increases quality, and decreases costs.
Smart manufacturing is often synonymous with Industry 4.0 and the fourth industrial revolution. These concepts represent a once-in-a-generation value opportunity that can deliver significant improvements in cost savings, capacity, and asset efficiency in manufacturing.
Smart factories represent a highly connected manufacturing environment that uses advanced technologies like IIoT (industrial IoT), 3D printing, robotics, machine learning, and data analytics to help improve production, uptime, and, ultimately, the bottomline. Perhaps, at the core, what smart manufacturing means to most people is getting the right information to the right place at the right time.
According to data from PWC, many manufacturers are onboard with IoT. Most (71%) said they’re building or testing IoT solutions—60% are leveraging IoT on projects within their facilities, 57% are leveraging IoT with supply chain and other partners, 42% with end consumers, and 58% with their business customers. Manufacturers are applying these technologies to solve problems in logistics, supply chain, and employee and customer operations, and they’re optimistic about it. PWC’s data said 93% of manufacturers believe the benefits outweigh the risks.
But manufacturers often struggle to bring an innovative IoT idea out of the PoC (proof-of-concept) stage. Instead, they get stuck in “pilot purgatory.” Manufacturing-specific hurdles can make it difficult to scale IoT solutions and applications, creating islands of innovation. How can manufacturers overcome these hurdles? As manufacturing marches toward a new era of connectivity, how will intersecting trends and developments like 5G affect smart manufacturing’s trajectory?
John Dyck, CEO at CESMII - The Smart Manufacturing Institute, says getting out of the PoC stage is a huge challenge for modern manufacturers pursuing digital transformation via the IoT. “We would argue that most of these projects struggle to get out of this stage,” he says. “It’s what we call ‘pilot purgatory,’ and it exists because the core data infrastructure layer required to collect, contextualize, store, and present manufacturing data to systems and solutions needs to be built from scratch for every project. There is virtually no reusability at this data layer, which means there’s little chance to prove the value of an implementation and replicate it broadly. Therefore, pilots don’t scale to larger implementations, or multi-site implementations stall after only a few sites.”
According to Wei Yuan, Industrial Innovation Lab manager at Hitachi America Research and Development, in many situations, a solution ends up being too specific to a problem and doesn’t generalize well, or it needs considerable modification during deployment elsewhere in the industry.
“For some factories and industries, there needs to be a significant investment to overhaul the existing low-efficiency age-old machinery with state-of-art machinery to enable smart manufacturing capabilities,” Yuan says. “Also, appropriate infrastructure and expert staff who can handle development and deployment of an industry-wide end-to-end (machine-to-user) architecture is fundamental to the conversion of smart manufacturing PoC to smart manufacturing application.”
Yuan says the IoT as a concept is often easy to understand, but defining or measuring actual ROI (return on investment) early in the process can be difficult, particularly in established manufacturing environments as there isn’t hard data as to total system implementation cost requirements. “With vague costing comes reluctance from executive teams to commit towards expanded implementation, testing, and validation,” Yuan adds.
Where industrial manufacturers see IoT's benefits
Q: Which of the following activities, if any, have you done and/or are planning to do in the next 2 years in relation to IoT?
Source: PwC 2019 IoT Survey
Another challenge in scaling IoT solutions is balancing scalability and effectiveness. “IoT solution providers are prone to general, scalable, and flexible IoT solutions, which could solve some common problems facing manufacturing industries, such as visibility to realtime shop floor status for agile response, customer demand, and inventory management for dynamic production scheduling,” Yuan explains. “However, to solve specific and more value-added critical challenges that manufacturers are facing, such as a production quality and production line/equipment efficiency, which are key for manufacturers to stay in business and keep competitive, specific operation know-how for manufacturing processes or shop floor machines need to be embedded into the IoT solutions to make them useful and effective to address critical production issues.”
Yuan says transferring such production know-how and digitized manufacturing knowledge through IoT solutions is much needed, but this can be challenging during solution scaling, since the production environment could change from product to product and from factory to factory. And yet, the IoT is imperative to achieve scale and to deliver the flexibility to normalize variable environments that each plant presents.
Scott McCarley, senior director of market development, connected operations at PTC, says manufacturers frequently struggle to get out of the PoC stage and remain stuck in pilot purgatory due to issues across three areas: financial, people, and operations. “From a financial perspective, pilots are typically technology-led rather than value-led.
As a result, there is a lack of clarity to financial impact and the ROI of pilots,” McCarley says. “From a ‘people’ perspective, the technology-led proof-of-concept approach fails to plan adequately for necessary resources or to address the digital skills gap and does not improve the work environment. Operationally, there is a lack of focus on high value, repeatable use cases, and there is often a lack of change management and user adoption planning across the enterprise.”
When solving for maximum impact, manufacturers must resource and plan to simultaneously roll out use cases across multiple sites and continuously reduce the deployment time in subsequent sites. McCarley says a traditional, sequential deployment process that deployed across 10 lines or even one factory every 3-6 months, delivering a 50-year approach to the global enterprise digital transformation is a plan designed for failure. “To deploy digital capabilities across tens of plants simultaneously, manufacturers need a strong foundation to scale and accelerate the journey,” he explains. “A solid foundation to enable scale consists of the right team, governance structure, and digital backbone to quickly iterate and achieve high-value impact across the enterprise within 24-36 months. The right team necessitates identifying dedicated people and ecosystem partners.”
To avoid pilot purgatory, it is important manufacturers get their ducks in order right from the get-go. For Nathan Hartman, professor of advanced manufacturing and director of the Digital Enterprise Center at Purdue Polytechnic Institute, this means being aware of and avoiding common pitfalls like a lack of executive support, an ill-defined vision, and a poor definition of project scope. “One thing I have seen too often is companies think they should somehow be ‘doing IoT’, but they have no idea why they want to do it,” Hartman says. “And to make the situation worse, the IoT technologies are often layered on top of broken processes and ill-conceived project and organizational structures.”
For manufacturers looking to scale solutions, there are certainly stop-you-in-your-tracks technology-related challenges in collecting, contextualizing, and storing manufacturing data. Even great solutions that prove value as a point solution often face significant challenges proving their value in different environments. Thankfully, there are some steps manufacturers can take to prevent this phenomenon—a phenomenon that John Dyck at CESMII calls “islands of innovation.”
Remote vs. Working Home
We would all agree we are living in uncertain times. The COVID-19 coronavirus is changing the world of work. The way we play. The way we live. On Tuesday, I hosted a radio show that was also unprecedented.
AI Against AI
What comes to mind when you think of deepfakes? A report by CB Insights got me thinking the other day about deepfakes and their impact on AI (artificial intelligence), quantum, and more.
Manufacturing IoT Revolution of Innovation
This should come as no surprise—since it has been publicized many times—the Chinese government has launched a very aggressive campaign to be a world leader in high-tech manufacturing by 2025.
Manufacturers Tighten Supply Chain
All weekend long I debated whether I should add my two cents about the sudden uneasiness over whether the COVID-19 coronavirus is really a wakeup call to us all and understanding the importance of the IoT (Internet of Things).
Crossing the Valley of Death
The IoT is a relatively new aspect of advanced manufacturing and, as with all emerging technologies, access to resources is critical. This is especially true for startup and small manufacturers. Michael Molnar, director of the Advanced Manufacturing National Program Office at the NIST (National Institute of Standards and Technology), says when a worthwhile and innovative concept fails to make it into production, he and his colleagues call it succumbing to the Valley of Death.
“There are a variety of reasons why novel ideas and inventions don’t make it across this valley,” Molnar says. “The progression from ideation to deployment is a complex process with many different challenges, and it is a rare technology transfer and development that is a straight or linear path on the TRL (technology readiness level) scale of maturity. Often, technology development could be plotted as iterative cycles, and the difficulty in negotiating this path lies in the disparate skills required to overcome these challenges. The majority of time, spanning the Valley of Death requires a team of the right people in the right places with the right skills working together to bridge the gap by fostering the development through the various stages. Occasionally, the stars align and this collates naturally, but, as the name ‘the Valley of Death’ implies, this is not often.”
But this is where having access to resources, like Manufacturing USA institutes, can play an important role. Molnar says each institute is carefully positioned and aligned within its specific advanced manufacturing technical space to provide the infrastructure, resources, and experienced knowledgebase to support and guide the movement of new innovations and inventions to bridge the gap.
“Another challenge inherent in many emerging technologies, such as IoT, is that the new technology fosters a new set of skills and players in this field,” adds Molnar. “Providing these individuals and teams with tribal knowledge concerning best practices and common pitfalls is critical during the scale-up process. As the scale-up progresses, workforce development is also key. The production of new technologies requires workers, engineers, managers, and others all to be conversant in the principles surrounding the new technology and (to) understand how to use it for the benefit of the business. Manufacturing USA institutes provide the resources for member organizations to collaborate to scale manufacturing technology and train the existing and future workforce in using these new technologies.”
Perhaps the most important step manufacturers can take, according to Andrew Dugenske, director of the Factory Information Systems Center and principal research engineer at the Georgia Institute of Technology’s Manufacturing Institute, is to develop an IoT strategy that relates to performance and profitability.
As the IoT powers a profound transformation in manufacturing, Sam George, corporate vice president of Azure IoT at Microsoft, says his team has learned some lessons. “We’ve learned that struggling proof-of-concepts are often related to undefined success metrics and unclear project ownership,” George explains. Microsoft and its partners work with manufacturers to implement best practices for avoiding struggling projects, including making sure manufacturers have a clearly defined business case with ROI goals and checkpoints to validate them. George says: “As an example, we think of the first phase of a project as the ‘proof of value’ phase versus ‘proof of concept’.”
Second, manufacturers need a clear understanding of modern software practices, like DevOps and CI/CD (continuous integration/continuous deployments). George says manufacturers also need to ensure that OT and IT are both involved in the design, deployment, and operation of manufacturing solutions; that there’s clear executive leadership support with empower teams on the ground; and that there are early discussions on “what’s next” after the proof of value has been completed successfully.
“Manufacturing environments are typically complex, involving different machines, processes, and existing software solutions,” George says. “These heterogeneous environments have created a lot of challenges in the past, ranging from missing interoperability between systems to data silos, in which data from one solution was trapped in databases and customers being unable to compare with other processes. The great news is that Microsoft and our partners have eliminated the technical challenges of scaling an IoT solution for manufacturing, leading with an open and interoperable approach. But of course, technology is only one aspect and, today, we often see organizational challenges that stand in the way and help (by) sharing best practices.”
IoT in Manufacturing Market Outlook: 2023
The IoT in manufacturing market was valued at $424 billion in 2016 and is expected to reach $994 billion by 2023, CAGR of 13.1% from 2017-23.
Source: Allied Market Research, 2019 Market Research Report
“Rather than developing a strategy that collapses under its own weight, an efficient plan should be developed that involves the various departments in the organization,” Dugenske says. “An important first consideration is how using the IoT can save money, enhance revenue, increase throughput, and improve quality.”
For most companies, Dugenske says developing an IoT strategy should be considered a process-improvement project versus a technology implementation plan. “Instead of thinking in general terms about implementing the IoT, companies should identify specific goals for using IoT,” he explains. “For example, if reducing the amount of scrap is important, implement IoT to collect the necessary data and determine what technological improvements are needed to meet the goal. Companies should avoid searching for a ‘silver bullet’ that will implement the IoT, and rather articulate their high-level goal. Simply, the goals should allow the organization to collect the appropriate data, turn it into information, act upon the information, and measure the results.”
The Future Is 5G
5G adoption in the manufacturing sector is still in the early stages, but more manufacturing companies are looking at how they can test 5G technology within their company today while planning for tomorrow. David Van Dorselaer, general manager of manufacturing and transportation at AT&T Business, says the key benefits of 5G for manufacturing include high bandwidth, ultra-fast speed, and lower latency.
“All these 5G features combined with other technologies, such as the Internet of Things, edge computing, machine learning, and cloud computing is where you really start to see the next level of innovation for the manufacturing industry,” Van Dorselaer says. “For example, manufacturing companies typically collect data from thousands of tools and equipment on the shop floor. Adding 5G+ in the mix enhances the scale and volume of the data being collected and improves the ability to process that data from many devices in near realtime. Manufacturers that can capture and crunch this data quickly will be able to produce actionable intelligence that can increase productivity and efficiency.”
Eventually, 5G in manufacturing will enable fully mobile, AR (augmented reality) training experiences where immersive and informative content is delivered just-in-time to help enhance learning, reduce errors, and improve information sharing. Van Dorselaer says, for example, a level-one technician onsite could have an engineer at headquarters guide him or her through the repair process remotely via 5G networks using context-sensitive 3D animations.
Galem Kayo, product manager at Canonical, says 5G eliminates the need for cabling infrastructure in a factory. As a result, a new class of mobile production assets will be introduced and wirelessly connected to factory networks. “The result will be a transformation of factory internal logistics with autonomous mobile units that will optimize work-in-process inventory management,” Kayo explains. “5G will drive the automation of internal factory logistics, leading to a reduction in working capital stuck in work-in-process inventory.”
Looking ahead, 5G, can help drive the future of smart manufacturing by unlocking new business models. “5G connectivity of production assets will unlock machine-as-a-service business models for machine builders,” Kayo adds. “The usage of production assets will be monitored and metered in realtime by OEMs (original-equipment manufacturers), allowing billing by the hour. Such business models will bring more flexibility to manufacturing, with a positive impact on capital productivity.”
As the era of 5G dawns, the future looks bright for smart manufacturing, despite the hurdles that often strand manufacturers in pilot purgatory and the Valley of Death. Georgia Tech’s Dugenske says the IoT has the potential to dramatically change the paradigm in manufacturing. Low-cost information technology, software, and hardware building blocks are allowing companies to easily collect, store, and process vast quantities of data that can be used to improve productivity. In order to harness the data and convert it into information that is actionable, manufacturers need to pursue a carefully articulated strategy that leads directly to decreased costs, faster time to market, and higher quality.
“Large realizations won’t come through serendipity; rather, deliberate and directed investments must be made now to take advantage of all that IoT has to offer,” Dugenske says. “Companies that are capable of developing and implementing an effective IoT strategy can reap great benefits, while companies that are not able to fulfill a strategy will find it increasingly difficult to compete.”
Want to tweet about this article? Use hashtags #IoT #InternetofThings #AI #artificialintelligence #5G #cloud #edgecomputing #machinelearning #bigdata #digitaltransformation #cybersecurity #blockchain #sustainability #futureofwork #manufacturing #IIoT
Staying Safe in the Surveillance Economy
Steve O’Brien, president of staffing, Job.com, talks with Peggy Smedley about the surveillance economy and how our online data creates value and is used as a tradable commodity. They also explore the regulation of the economy, how it is taxed, and the responsibility of the companies collecting the data in terms of privacy and transparency.