Robotic Process Automation (RPA) has been a useful tool for many organizations. Despite the initial fear that it would grow to take over the jobs of developers, many have come to see that RPA and automation only function well when they work in tandem with developers. 

According to Yishai Beeri, growth technologies lead at LinearB, the best way for organizations to utilize RPA is to implement it with the purpose of eliminating the mundane tasks that would usually fall to developers. 

He also explained how this technology works to ensure consistency across a development team. Beeri said, “Developers have their own skills, they can automate basically whatever they like if they put their time into it, but sometimes, you want a more organized or central solution for automating these things instead of every developer just scripting away,” he said. “Maybe it’s not important enough for a single developer but if you look at 100 developers…the small time wasters are things that you can automate away with a more centralized solution.” 

Carlos Melendez, COO and co-founder of Wovenware, echoed Beeri’s sentiments by explaining that organizations would much rather have their developers working on tasks that bring value to the company rather than spending the majority of their time on duties that could easily be automated with something like RPA. 

Melendez also explained that when implementing RPA, this is the message that can fight off the employee resistance that may come from the fear of losing their jobs to automation technology. “A lot of the time it’s not about replacing employees, it’s about augmenting their capabilities. So, if half of your time is spent kind of preparing a file or preparing an integration or moving data from one point to another or doing data entry, then you want that person to spend more time on their analysis and verifying what is happening instead of the actual data entry part,” he said. 

RPA still new, and evolving

Jon Knisley, principal consultant, automation, and process excellence at FortressIQ, said that RPA and other automation technologies are still relatively new and, therefore, rapidly evolving. He said, “Among companies that have deployed RPA, a majority have less than 10 bots in production and just 10% have launched more than 100 bots, according to a recent report from Automation Anywhere.” With this, he added that he believes that the full breadth of what RPA and automation in general can do is still undiscovered. 

“Only 11% of business executives surveyed by McKinsey believe their current business models will be economically viable through 2023. Given the potential disruption, organizations continue to invest in complex change programs despite dismal success rates of less than 30%. Automation is the new transformation,” Knisley said. He also noted that RPA has been the fastest growing segment for the enterprise software market for three consecutive years beginning in 2019. “Grand View Research estimates the global market for RPA will surpass $2 billion in 2022 and continue to grow annually at 40%,” he said. 

Arthur Villa, an analyst at Gartner, said that his company’s research has yielded the same results, saying that he has seen no evidence that RPA has been slowing down, even in the midst of newer technology. “[As far as] the state of RPA implementation, I would say that it is still in the relatively early days. If we look at it as a four quarter game we’re probably only in the second quarter… I think that there’s still a lot of adaptations that have been made in the last couple of years,” he said.

With this, Villa pointed out that RPA has only been growing in popularity as larger and more well-known organizations introduce this technology. “A lot of these new vendors are coming into the RPA market and shaking things up. There’s a lot happening within the market especially from the customer and buyer perspective… Many companies start small with RPA and then they rapidly expand those programs so I think we’re still early on in new customers buying RPA and beginning to experiment with the technology,” he said. 

Villa believes that the reason RPA has been so widely accepted and implemented is because it offers organizations simplicity and overall convenience. According to Villa, when compared to other AI solutions, RPA is lower in cost, easy to understand, and companies will usually see a quick return on investment. 

Data, change management challenges

In spite of this, though, it is not uncommon for organizations to face some challenges when introducing RPA into their business processes. According to Melendez, these challenges often fall into two categories: data driven and change management. Melendez explained that the data aspect of these challenges has to do with the quality of the data itself as well as overcoming the different types of roadblocks that arise when you try to automate using bad data. The change management sentric challenges have more to do with employees being worried about what RPA is going to do and how it will change their own jobs. 

When working to remediate these challenges, Melendez said, “Technology is moving so fast that you really need a good set of technology partners that you can trust, that you can go to when you need certain technology solutions. You have your AI partner, your RPA partner, and other partners that will help you navigate the complexities and the changes in those technologies.”

From RPA to IPA (not the pale ale)

The conversation around RPA has shifted slightly in recent years in order to cast a wider net, the newer terminology is Intelligent Process Automation (IPA). The low-code automation company Nintex has been championing it, and according to Terry Simpson, senior solutions engineer at Nintex, “IPA is like the grownup and more mature sibling of RPA. When we say sibling, think about IPA being about 20 years older than its younger sibling RPA, on the maturity scale.”

Simpson continued, “IPA is actually the combination of several technologies coming together to create a very mature and flexible automation capability. Intelligent workflows, natural language processing, machine learning, and even RPA are all integral parts of an IPA solution.” He explained that a key difference between RPA and IPA is that while RPA usually runs on a local machine, IPA is a cloud-based virtual environment. “In simple terms, think about IPA sitting right in the middle of all your applications and performing process automation focused on an entire solution, not just tasks. Tasks may make up a piece of the solution, but IPA brings the entire solution or process together,” he said. 

Brett Greenstein, data and analytics partner at PwC said, “RPA is getting less discussion… because automation has expanded well beyond screen scraping and bots, through the use of APIs, Microservices, and AI/ML. Many companies have adapted to this by expanding the term to IPA to include those newer capabilities as well as process mining.”

Greenstein explained that in the current environment the need for automation is only growing. In the midst of the great resignation and a shortage of skilled developers, automating tasks using a smarter solution is quickly becoming a necessity rather than a luxury. This increased demand for automation has led to the expanding of RPA into IPA in order to introduce fresher technologies into an already reliable method of automation. 

RPA as communication tool

Beeri thinks there is a new role that RPA can fill in the face of a more distributed workforce. He said that RPA can be used as a communication tool that can remind developers of when it is time to take the next step. “A lot of the work that software developers do as a team is a lot of back and forth and communication between people… so coordinating that in an environment where it is mostly asynchronous and we’re not in the same room anymore… automation and smart bots can really help in coordinating this ‘dance’ between people so that people are not interrupted,” he said.

Beeri said that even though this is not a task that has been done in the past or a role that used to be filled by a separate employee, it has become important with the trend we’re seeing towards working remotely. He said, “It really helps to minimize interruptions and maximize speed when working together on things.”

According to Villa, only a few organizations are currently experiencing the full benefits of RPA that Beeri is referring to. He believes that the majority of companies utilizing this technology are the ones that are generating high volumes of revenue, meaning that small and mid-market organizations have yet to adopt automation technology. He said, “There’s still a lot of education that has to happen within mid-market companies that need to understand ‘what is RPA? How can it be used? And how can I get the most bang for my buck?’”  

Knisley also pointed out that education around RPA and automation is essential when trying to implement it in the most effective way possible. However, he also placed an emphasis on the importance of fully understanding and optimizing the company itself before introducing automation. He said, “To achieve the magical future state promised by technology, companies first need to understand their current state. Unfortunately, most companies do not understand how they truly operate especially at a gradual user activity level. To be successful and avoid false starts, companies need to discover, re-engineer and automate — in that order.” 

RPA has fallen out of the spotlight over the past few years

Even with RPA’s rapid growth, it is noteworthy that it has fallen out of the spotlight somewhat in recent years. According to Geenstein this is the result of newer technologies being introduced. He said, “First, there is the screen scraping and click automation that allows RPA to execute the same steps a person would execute in any application. Second, there is the scripting for bots that identifies a sequence of actions with basic logic to decide what action to execute next. As APIs and microservices become more and more available, especially as applications modernize on the cloud, the need for screen scraping and click automation goes away.”

Along the same lines, Beeri said that he feels RPA is not as widely talked about because, at its inception, it was overhyped and now it is failing to live up to all of the original promises. “I think when you start to look at how to deploy [RPA], and what tasks need to be removed, you’re finding that you can change the actual task, you don’t stop at just putting a robot in to automate data entry… The solution for the problem at that point might not be RPA, it might just be automating something using no-code or low-code methods,” he said. 

However, Melendez credits this lack of discussion to something different. He said that rather than RPA being at the center of the discussion, people have shifted to speaking more generally about automation. “RPA as well as AI is becoming so prevalent that the conversation is no longer about deploying RPA, it’s about the solution that we are going to deploy [using RPA],” he said. Melendez explained that because these tools and technologies are so advanced, not only is it assumed that they will be in place, but that it is also assumed that, in most cases, they are going to be able to easily automate whatever is necessary. 

Automation and humans need to work together

On an SD Times-led discussion of RPA on the Discord “Dev Interrupted” server, participants had a lot to say. One of the respondents, Dr. Don Wilcox, talked about a robot used at his organization, which they named Marvin. “We have automation that completes a Task (the most-specific sort of ticket) when a PR associated with that Task is completed,” Wilcox explained. “Then Marvin takes over and changes the state of the parent story based on whether the dev tasks, qa tasks, demo tasks, etc, are complete. Marvin has his own row on the board, and you can get him to perform certain automation tasks (such as adding the standard stories and tasks to a new sprint) just by giving him an appropriately named task in that sprint.” 

This is a good example of the way that automation and human employees have to work together. There’s no doubt that the addition of the robot makes things run more smoothly but the robot cannot function without the direction of the human, which was the overall consensus from the discussion. The idea that RPA or another form of automation will be the end of human labor is far from the truth.

In further support of this, Wilcox said, “Once you build a robot, someone needs to maintain, enhance, QA, replace. That robot assumes it’s own product lifecycle, which will likely require humans. For the foreseeable future, it is going to be humans building the robots, even if the robots help.”