Robotics | Article

What Humans Need to Know to Drive Automation

What to do (and what not to do) on the journey into Industry 4.0.

Written by: Poornima Apte

While manufacturing and logistics are no strangers to automation, the advent of Industry 4.0—with the concept of a smart factory running on big data and machine learning—is poised to accelerate its rate of adoption.

According to research firm MarketsandMarkets, smart manufacturing is expected to track at a compound annual rate of 19.7%, increasing from $4.4 billion in 2019 to $10.8 billion by 2024. Given that high-voltage automation is going to be a part of tomorrow’s business landscape, understanding and embracing the elements that work for your enterprise is key.

The right fits

Enterprises lean on automation depending on the scale of production, diversity of manufactured goods and supply chain logistics. Decision-makers usually pick from one or more of the following three broad approaches:

The first is fixed automation, where equipment is leveraged to meet industrial automation goals. While incredibly effective at its set task, fixed automation can be inflexible and incur steep costs. Jason Barton, chief commercial officer at Realtime Robotics, says large-volume production in automotive manufacturing meets this definition. “Such automation is best used for long runs [and] high production rates, such as the BMW 3 series where you know the model line is going to be running for the next three years,” he says.

Second, programmable automation refers to technology where the specific tasks to be automated for each batch of goods are programmed in sequence. Relying on such automation means workers can simply change the programs used to produce parts—for example, a robot changing its end gripper—without having to shut down the entire production line. This delivers an increase in key performance indicators such as manufacturing uptime.

Finally, flexible automation is an extension of programmable automation, except the various changeovers needed to produce different products are programmed off-location. Much more nimble than programmable automation, production need not be relegated to batch mode. Barton offers the example of collaborative robots, which work alongside human workers, as an example of such automation.

As companies progress along their path to automation, machine learning algorithms can kick in to deliver predictive maintenance for entire fleets of production machinery.

Use automation like a fine scalpel

Robotics usually is not a direct one-to-one robot-to-human replacement for all processes, says T.J. Johnson, former chief technology officer at two robotics startups and currently president at TJ Johnson Robotics & Consulting working in general integration, robotics system design and workforce integration.

Meanwhile, Barton says that in most instances where automation is pressed into service, the actual robots only make up a quarter of the final tab; the related integration and programming costs are much more steep. “Medium-sized manufacturers have found it prohibitive to really embrace automation,” he says.

To arrive at a better ROI, use the kind of automation that makes sense within your context. For many, that might mean flexible automation, which has a lower integration costs because of its adaptability and ease of use.

Do not go for the complicated first

Start by reaching for the low-hanging fruit, Barton says. “Do not try and automate everything at once. Target relatively simple processes, then slowly expand automation to more complex processes within the manufacturing lifecycle.”

Mobile tablets on the manufacturing floor can pull up forms to be signed off on electronically or send automatic inventory updates, increasing efficiencies and decreasing room for manual errors.

Also, as companies progress along their path to automation, machine learning algorithms can kick in to deliver predictive maintenance for entire fleets of production machinery. Industrial Internet of Things (IIoT)-embedded sensors can read performance metrics and automatically shut machines down if their indicators exceed certain pre-set numbers. A machine learning algorithm can detect if rotors heat up beyond a certain temperature profile, for example, and notify operators through a text alert. Such predictive maintenance increases valuable manufacturing uptime.

Do not accelerate too quickly

There is no need to go in all with all cylinders firing. Allow ample time so any early wrinkles in the process can be ironed out before ramping up implementation.

“First, decide what the gold standard looks like at your company before going all in,” Johnson says. “For some companies that gold standard can mean flexibility and optimization first, while for others it can mean high volumes of production is prioritized.” A tactful approach testing the waters is always best, he says.

Do not leave your people behind

Johnson says companies should be open about the process with their employees. The misconception that robots are set to take jobs away has deep roots in the fear of the unknown, so pointing out that automation can actually help employees do their jobs better is one way to assuage fears and get everyone on the same page.

“If you bring your employees in early, you can make it a positive interaction,” he says.

He adds that actively engaging employees early on in the process leads to deeper conversations about digital transformation and how managers can help everyone fit into the new landscape, whether it is through new skillset training or additional education.

“The question of digital transformation is really about, ‘what is the future and what is my role in it?’” he says. “So, if you care about your employees and show that you care about them and that they have a role in your collective future, they will help you succeed.”