Moving from automated to autonomous manufacturing
When most of us pick up a candy bar, we briefly anticipate enjoying flavors like chocolate, caramel, and nuts all swirled together into a tasty treat. Few ever contemplate the exacting science that goes into assuring those sweets are consistently great every time we unwrap them.
Yet despite the highly automated nature of most modern factories, they don't always roll off the conveyor belts exactly as intended. Sometimes the ingredient mix is a little off. Other times, there might be a packaging mistake. Once in a while, the finished confections are slightly heavier or lighter than mandated weight regulations. In your kitchen, none of these issues would be a big deal. But when candy makers discover such errors, they have to determine whether it's financially viable to do something about it. Do they suspend processing? Recalibrate machinery to the correct formula? Throw thousands of candy bars away?
Even with a certain margin of error, automation allows manufacturers to produce millions of quality products each year. But every manufacturer wants to improve overall equipment effectiveness and quality to avoid the types of inefficiency, waste, and shutdowns that often result in multimillion-dollar hits to the bottom line.
That's why many manufacturers these days, even if they are not aware they are doing so, are beginning to think about evolving from automated to autonomous systems.
Is it automated or autonomous?
The difference between the two approaches is a thin one. But think about automated as being a somewhat fixed process where production runs pretty much on its own unless interrupted by human intervention, a power failure, or acts of mother nature. You might monitor what's traveling down the line to identify potential maintenance or production issues, and you may even have futuristic robots performing tasks in cages, controlled by applications sitting in a centralized data center. But it's all rather static, pre-programmed, delineated, one-dimensional—like automobile assembly lines of old that produced one or two car models with little variation.
Autonomous systems, on the other hand, begin to make use of Industry 4.0 technologies—the Industrial Internet of Things (IIoT), along with artificial intelligence (AI) and data and analytics—to adjust and optimize production on the fly and even enable customization at the point of manufacture. In fact, compute power moves out of the data center to the industrial edge, near machines on factory floors.
Once there, they are better able to collect, integrate, analyze, share, and act upon data from multiple assets and supply chain sources. So, if a machine pouring chocolate into a candy bar mold suddenly starts adding 10 grams instead of the prescribed 9 grams, the proportions could be quickly and even automatically rebalanced without much, if any, human intervention. When a potential error is detected, the problem's source is identified. AI or some algorithm determines the necessary adjustment and the issue is autonomously corrected, sometimes even before it becomes an issue.
As these types of predictive capabilities become more prevalent, manufacturers are likely to view them as important for maintaining quality control, limiting supply chain costs, reducing waste, and improving their environmental sustainability postures. What's more, they could also help ensure business continuity if another global pandemic or a similar crisis forces them to send workers home once again.
Please read: Object detection technology: The next big thing
"The purpose of autonomous systems is to do more with less," says Rian Whitton, a senior analyst at ABI Research in London. "The more machines you have that can work independently, and the more use cases and applications they can perform without human intervention, the easier it is to redeploy workers to address current and future needs. This is vital for companies trying to manage through tightening labor markets in aging societies."
In the not-so-distant future, it's possible to imagine self-sustaining smart factories with interconnected machinery that communicates, sharing information and looking out for the well-being of the total production environment. Robots will bust out of their limited programming box and be able to both learn and perform sometimes complex tasks, relieving humans for other critical work. Enterprise data will integrate with supply chain and real-time production data allowing for better informed business decisions. Digital intelligence will exist everywhere, from the edge to the cloud. Security will be indispensable. Consumption-based models will dominate. Everything will function to achieve certain manufacturing efficiencies and outcomes. Moreover, processes will self-manage and self-service.
Please read the report: Building tomorrow's enterprise
Ralf Hagen, engineering manager at Nestlé Deutschland, is driving a transition from automated to autonomous systems for several of the multinational food company's factories in Germany. He says a key advantage of autonomous systems is that they handle security and maintenance behind the scenes.
"We are surely not going to have the number of skilled systems administrators in our company to do this because we are a production company," he says. "We are not an IT company. We are not an infrastructure company. We are just a simple producer of food. We want systems to be like a smartphone: Everyone has one in their pocket but doesn't have to think about how it works—it just does. On the industrial side, we can learn from this consumer experience."
That's the goal of going from automated to autonomous. The concept is built on a solid foundation of people (skills, processes, and culture), technology (an edge-to-cloud industrial digital foundation and data pipeline), and economics (investment strategies and flexible financial and business models for accelerating growth). And if you can achieve the right balance of those three dynamics in your smart factory, it can lead to major improvements in productivity.
Indeed, in a 2019 Deloitte survey, more than 86 percent of respondents said they believe smart factory initiatives will be the main driver of manufacturing competitiveness in the next five years, which could potentially ignite stalled productivity growth as businesses begin recovering from the pandemic.
"Smart factory technologies have the potential to dramatically accelerate recovery," says Paul Wellener, vice chairman and U.S. industrial products and construction leader for Deloitte LLP. "Manufacturers who were already piloting technologies such as analytics, sensors, and wearables will be able to make their workplace safe and resume production faster."
Wellener says many companies will be trimming budgets in the coming months to gird against economic difficulties. But he doesn't believe that will extend much to automated or autonomous technologies.
"I expect these initiatives to accelerate as manufacturers deal with workforce limitations," he says. "They know a shift toward digital tools and technologies can help maintain optimum production levels and pivot processes to become more agile and responsive in the face of any future crisis."
But even as such initiatives accelerate, industry watchers acknowledge manufacturers still have a long road to travel before their factories can claim to be running on mostly autonomous systems.
Beyond traditional data centers
That's because, despite advances over the past few years, most of them still rely on older automation where data and apps aren't easily shared across machines or production plants; where planning and production tend to be more linear and limited than flexible and futuristic; where security architecture is likely built around decades-old operating systems instead of modern cloud or IoT services; and where machines aren't easily accessed or managed from afar—they weren't built for that.
Please read: Factories and the challenge of wireless networks
Nestlé's Hagen says moving from legacy architectures to more efficient digital autonomous technologies has been a multi-year journey for his company. But with COVID-19 requiring new approaches to production and remote work, these efforts have accelerated in the past several months.
"With the pandemic, we have been told that delivering this [autonomous technology] is not just urgent, it's over-urgent and overdue," he says. "A lot of things have been made possible the last six or seven months that would have taken two or more years to accomplish previously. Hopefully, we can maintain that rate of change."
For most manufacturers, the ability to shift to autonomous systems will ultimately require aggregating, normalizing, and analyzing huge volumes of production data from numerous sources in the enterprise and beyond. According to Gartner, only 10 percent of enterprise-generated data was created and processed outside of traditional centralized data centers or clouds in 2018. But by 2025, Gartner expects that figure to reach 50 percent as computing power becomes more decentralized and Industry 4.0 projects proliferate.
The key to successfully using that data to improve operations and induce better business decisions will be to collect, aggregate, and cleanse it—then to let it flow freely across enterprise organizations.
According to McKinsey Global Institute, there is an overall positive correlation between each type of data flow and GDP growth. McKinsey estimates that global flows contribute between $250 billion and $450 billion of growth every year to GDP. At a micro level, one can imagine how letting data flow within individual enterprise organizations might contribute to their bottom lines.
The necessary evolution of manufacturing from automated to autonomous models is well under way. Major enterprises have been investing in technologies to help manage and make sense of all the data within their purviews. As that continues and emerging technologies such as 5G, AI, virtual reality, augmented reality, and predictive analytics mature, expect to hear more about autonomous systems.
We want systems to be like a smartphone: Everyone has one in their pocket but doesn't have to think about how it works—it just does. On the industrial side, we can learn from this consumer experience.
This article/content was written by the individual writer identified and does not necessarily reflect the view of Hewlett Packard Enterprise Company.