Automating behind the scenes: Robotic process automation for back-end support
Shop floor automation helps manufacturers eliminate repetitive physical tasks, freeing humans to focus on value-added work. Robotic Process Automation extends this approach to the back-end processes that support the shop floor.
June 21, 2022 by Jacob Stoller
One of the most important applications of industry 4.0 technology is the early detection of deviations that could lead to inefficiencies, product defects, equipment failure, safety hazards or other problems. The value of these warnings, however, depends on the processes that ensure that appropriate interventions take place.
The challenge for many manufacturers is that many of their data-handling processes are manual, creating a weak link in the chain between important signals and timely response. The gap is driving the adoption of what has been dubbed Robotic Process Automation (RPA), which is essentially the deployment of sequences of computer code, called bots, that execute repetitive tasks in robot-like fashion.
“It [RPA] provides value in automating processes based on structured data, many of which previously required manual intervention,” says Dinesh Nirmal, general manager at IBM Automation. “For manufacturers, RPA bots can be used as input to manufacturing analytics tools to analyze the overall progress, identify gaps and find improvement opportunities. More specifically, using RPA can help businesses improve go-to-market time by eliminating routine tasks that would have required human labour.”
“The essence of manufacturing is mostly repeatability,” says Gavin Davidson, industry product manager at Oracle NetSuite. “You design a product and end up with a list of materials (BOM) that are required along with a process to make it. At its core, RPA can assist manufacturers in repeating actual processes, or by ensuring that deviations are properly routed or handled.”
Repetitive mind-numbing work that’s prime territory for automation can often be found behind the scenes of existing shop-floor automation.
“Whenever there are robots on the shop floor, there’s going to be a host of processes that surround that, maybe from a maintenance perspective, a quality control perspective or a safety perspective,” says Sherief Ibrahim, GM, business applications, Microsoft Canada. “It’s one thing to say we’ve got a machine that can automate a task. But how well are we automating the human intervention or human processes surrounding that task?”
These repetitive interventions – alerting a supplier, issuing a work order, removing a safety threat, correcting a quality problem – often extend far beyond the shop floor. Davison explains that beyond the manufacturing process, there are many opportunities for RPA to automate other business processes such as engineering changes, approvals, vendor onboarding and performance monitoring, planning, inventory management, invoicing and collections.
The most common ways manufacturers are using RPA, according to Nirmal, is to automate some of the repetitive supply chain and inventory processes. He notes that global supply chains can be complex given that manufacturers may work with suppliers in multiple countries.
Recent technology developments have enabled RPA to take off. “There’s a lot more data out there, and there’s a lot more digitization,” says Ibrahim. “Every organization’s assets from machinery, cameras, thermometers and more are connected to the internet, giving us more signals than we’ve ever had.”
Another important improvement is better interoperability between diverse apps. For example, a maintenance workflow might touch on production scheduling, equipment support, and procurement. “That workflow can be automated because we’re now able to connect all the apps and machines that are involved,” says Ibrahim. Power Automate, a Microsoft tool used to design and implement RPA workflows, can work with 580 different connectors, he notes.
Software for designing and implementing RPA automation is also becoming more user-friendly. According to a recent case study referenced by Ibrahim, an operations analyst at Rockwell Automation used Power Automate to automate a workflow involving 150 file locations. This triggered a number of employee-initiated automation projects.
“People don’t think of RPA as something that a business user can handle,” says Ibrahim. “They think it’s complex and can only be handled by experts. You do need to have the app developers engaged to help govern how RPA is managed. However, we have more and more case studies where a person who’s not a developer at all was able to take the platform and automate a process.”
The incremental nature of RPA also allows for solutions that pinpoint specific problems. “The start of automation really comes from capturing and understanding data that is relevant to a process,” says Davidson. “You can then begin by focusing on that data and designing repeatable processes to review, analyze and act on the data in order to generate a positive outcome.”
RPA solutions, however, should not be seen as a substitute for comprehensive solutions that unify workflows across diverse departments and business units. “RPA or robotic “process” automation is a misnomer,” says Nirmal. “It should have been called robotic “task” automation. RPA can be useful to automate a task and includes a workflow automation, but it is not intended to be used to orchestrate work across multiple people or multiple systems.”
Security and governance can also be of concern when incremental solutions begin to show up on the network. “I don’t want someone to automate a process where something sensitive ends up posted on Twitter,” says Ibrahim. “So we need to have guard rails to prevent sensitive information from flowing into the wrong places. An organization’s IT or developer team has built-in governance tools that will make sure Robotic Process Automations stay compliant and follow a company’s policies.
It’s also important to keep in mind that automation has its limits. “As automation adoption increases, more human tasks are automated in today’s digital world,” says Nirmal, “but there is still a large portion of our daily work that requires manual labour, even if these tasks are repetitive. This is because some tasks still require human cognition and intuition to receive optimal results. RPA bots are programs and can make use of AI to help them to make sense of the world, but they cannot think by themselves beyond simple and well-defined tasks.”
The key is that RPA can’t solve every problem, nor is there a magic starting point.
“The best opportunities for using RPA will vary depending on the existing technology in the workplace and the desire to invest in or implement them,” says Davidson. “Investing in new machinery and robotics might offer a quick way to automate manufacturing processes, and utilizing your business systems automation capabilities would be another quick win. Usually, the best place to start is with a simple repeatable process of any kind that can be easily measured and monitored to ensure success.”
Jacob Stoller is a journalist and author who writes about Lean, information technology and finance.