When we imagine a manufacturing facility, we think of conveyor belts, robotic arms, elaborate machinery, and human workers in protective gear ensuring the entire system’s functioning correctly and putting the finishing touches on goods before shipment.
Automation is, simply put, built into our conception of how modern manufacturing operates. But recent technological advances have fundamentally changed what automation can accomplish, and what we once envisioned as the factory of the future is in many ways already here today.
This new era of automation will disrupt and displace old models of production and permanently change not only manufacturing, but product development, supply chain management, and sales and marketing. This post explains how we got here, what this new technology means for automation, and how forward-thinking manufacturers can apply it to their own operations.
A Brief History of Manufacturing Automation
Automation, in the sense of using machinery as a substitute for human labor, dates at least back to the Roman Empire, where water wheels were used extensively in mining projects to drain water from underground shafts.
The modern era of manufacturing automation was shaped during the Industrial Revolution in the 1800s when agrarian and artisanal modes of production were supplanted by industrial ones based around machine manufacturing. Henry Ford’s assembly line techniques in the 1910s and urban electrification in the 1920s furthered this transformation
In the 1940s, the Ford Motor Company established an “automation department”—marking the first use of that term in relation to manufacturing. During this period, feedback controllers were rapidly adopted in manufacturing to facilitate numerous production processes, including accurately positioning large objects (like partly finished automobile components) on a production line. It was against this background, in 1946, that Fortune magazine published a cover story titled “Machines Without Men” which sketched out a near-future vision of workerless factories.
The programmable logic controller (PLC)—a computer designed for usage in rough industrial settings—was invented in 1968. And in the 1970s, the distributed control system (DCS) was developed by a team of engineers at Honeywell. By the 1980s, computer-driven automation had been widely adopted across the manufacturing industry and was even being employed in sectors like retail and pharmaceutical.
How Automation is Being Leveraged Today
Automation initially aimed to mechanize certain repetitive physical tasks in the manufacturing process, and advances in robotics in the 1970s and 1980s helped amplify this transformation as industrial robotic arms were widely deployed for automotive manufacturing. But today’s automation is being driven by sophisticated technologies like artificial intelligence and machine learning that can mimic human thinking as well. This development offers several key benefits for manufacturers, namely:
- Harnessing digitization. AI and machine learning go hand-in-hand with the Industrial Internet of Things (IIoT), as computers can not only process data from digital sensors embedded in physical devices, but use that information to make real-time analytical decisions about your entire production process.
- Lowering costs. Automation allows you to strip complex production processes to their most basic functional components and then optimize those activities for increased efficiency and lower operating costs. Machine learning and AI can amplify this by identifying, for example, opportunities to reduce waste and optimize energy usage, and even predicting and forecasting demand in order to allocate resources more efficiently.
- Increasing productivity. The more automated your production process, the more efficient your operation. In fact, end-to-end factory automation can double or even triple production compared to more limited automation deployment, according to the Robotic Industries Association. AI can also help surface data-driven market insights that can guide your research & development, product design, and customer support.
- Increasing workplace safety. Dangerous and even potentially life-threatening manufacturing tasks are often the first ones delegated to robots, which are never distracted or fatigued. And Industrial IIoT provides real-time visibility into facility floors, which means managers can track worker activity, machinery compliance, and other relevant safety issues. Machine learning, meanwhile, can facilitate predictive maintenance models which help you avoid costly (and potentially harmful) equipment outages and safety shutdowns.
We can see how these factors are already disrupting manufacturing operations across sectors:
- Automotive. The combination of robotics and AI-powered data analytics is allowing automotive plants to deploy just-in-time production processes that can quickly respond to demand shifts, thus reducing waste and costs.
- Aviation. Aerospace manufacturing has always been specialized and complex, but recent developments in smart manufacturing have helped to simplify many phases of production, such as drilling and filling, by using a combination of robotics and human labor.
- Chemical. Automated batch manufacturing allows for rapid transitions between product categories (like different colors of paint), facilitates more effective quality control checks without human intervention, and maximizes production with a minimum amount of waste.
- Electronics. Robots and mechanical arms are increasingly being integrated into production lines to limit the number of repetitive moves workers need to make, increase precision (especially important with sophisticated electronics), and reduce errors.
- Pharmaceutical. The pharmaceutical industry is using robotics to automate complex processes like high-throughput screening, nuclear magnetic resonance, and high-performance liquid chromatography. AI is also being leveraged throughout the drug discovery and development process, from synthesis to testing to manufacturing.
Getting Automation Right: Key Considerations
Automation offers countless benefits for manufacturers, but like any other technology it should be deployed strategically—not as a quick fix for structural flaws in your production process. Before you invest, consider these five factors identified by McKinsey:
- Technical feasibility. It’s obviously important to understand which manufacturing activities can actually be automated using currently demonstrated technology. You may find that some of your procedures can be completely automated (like welding, soldering, and packaging), some partially automated (like drilling and filling), and others not automated at all (like driving supplies between facilities—although even that might be coming soon).
- Costs to automate. Automation offers substantial savings over the long-term, but those gains will require considerable investment upfront. It’s worth considering which automated processes will offer the most substantial ROI. One option for this type of assessment is profit-sensitive scoring (detailed in this white paper), which can be an effective way to evaluate the value of certain automation projects in terms of cold, hard cash.
- Worker trade-off. Automation facilitates both mechanical tasks and goal-oriented decision making, thus freeing up human labor for more value-added activities. But it’s important to consider the relative scarcity, skills, and costs of the workers who would normally be performing these tasks.
- Benefits beyond labor-cost substitution. As detailed above, automation can often improve product quality control, workplace safety, and other factors beyond simple cost of labor assessments. It’s important to convey these kinds of benefits to stakeholders.
- Regulatory and social-acceptance considerations. Anytime you’re installing mechanized equipment in a factory setting you need to carefully follow relevant guidelines, whether federal, state, or local. And you should always emphasize the ways in which automation will enhance, not supplant, human labor in your operations.
One key consideration not identified by McKinsey is choosing the right production system. Automated manufacturing typically involves one of three types of production systems:
- Fixed automation. Also known as “hard automation,” fixed automation refers to a production system where machinery is programmed to perform simple, repetitive tasks. Fixed automation is effective for producing high-volume, low variability parts or components.
- Programmable automation. This production system is often used for manufacturing products in batches (like the chemical manufacturing example above), whether several hundred or several thousand units at a time. Operation sequences are controlled by coded instructions which can be quickly changed to accommodate new configurations.
- Flexible automation. Flexible automation (or “soft automation”) systems are similar to programmable automation systems but provide even greater adaptability, as a variety of different parts or components can be produced with virtually no time loss as the system adjusts to new instructions. The machinery necessary for flexible automation is quite sophisticated—such as robotic arms that can perform multiple tasks—and represents a considerable initial investment.
The production system you install will reflect the products being produced, the machines required for implementation, and the available resources.
Manufacturing is at the forefront of a technological revolution driven not only by robotics, but also by the adoption of artificial intelligence, machine learning, and Industrial IIoT. Together, these technologies have transformed our conception of what’s possible with automation in manufacturing, and by successfully implementing them you can optimize your entire production process, deliver substantial ROI, and consistently outperform industry competitors.
Ready to learn more? Check out our Digital Manufacturer: A Blueprint to AI infographic, which identifies key automation opportunities and details how to become a fully developed digital manufacturer.
Machine learning has changed manufacturing forever, and for the better. Here are some of the ways ML will continue to revolutionize the industry in 2021 and beyond.
Learn exactly what industrial internet of things (IIoT) is, and why there’s an even better way forward—the intelligent industrial internet of things.
Get a complimentary copy of the 2020 Forrester Wave: Multimodal Predictive Analytics And Machine Learning Solutions