How low code is killing off RPA

Computer programming source code
(Image credit: Shutterstock/BEST-BACKGROUNDS)

As I know from my own experience developing the technology and then expanding its use across all kinds of sectors, robotic process automation (RPA) has scored some quick and valuable wins where optimising business processes is concerned. For example, one major commercial insurer has saved over 50,000-man hours through using RPA in its claims management.

While the benefits of RPA have been hard to ignore, growth in this technology is slowing down. The usefulness of RPA for automating existing individual processes is ending, along with fewer use cases as organizations move onto a bigger mission of enterprise-wide automation.

Currently, applications of RPA typically look to automate processes across existing legacy applications through the usage of bots. But in the intelligent automation era, these very apps and processes are being fully digitized which means there is less requirement for RPA bots.

Vision to create

Most legacy applications and processes have been created over the past decades to automate functions and back-end processes that computers would normally require manual input for - for example, account opening or signing up for a new insurance/phone plan. To optimize “computerized” legacy processes, RPA automates those processes without any changes to the underlying infrastructure or accessing any of the legacy code. But now any new process, built digitally first, is likely to be far more autonomous making it adaptable and connectable to anything else without needing RPA beyond temporarily plugging short-term integration gaps.

RPA can be a useful start to streamlining operations and automating tedious costly human centric tasks that are prone to errors, and which hold back employees from doing more valuable work. By comparison to how RPA patches suboptimal processes, it is low code tools that offer the strategic vision to create, optimize and transform critical business processes that drive hyper-automation and innovation.

Among the reasons why low code will assume a bigger role is how it speeds up the delivery of new, purpose-built processes, providing enterprise-grade applications and reusable assets that are always ready to further optimize and automate. Furthermore, low code tools and platforms suit the ever-growing need for closer collaboration between technology and business departments to achieve a broader, deeper digital transformation. This allows the cross-divide communication on how to transform as well as digitally automate processes end to end for the benefit of all.

Francis Carden
Francis Carden

Francis Carden is VP of Intelligent Automation and Robotics at Pegasystems.

For and against

As low code continues to eclipse RPA as the predominant automation technology, there will continue to be some arguments, twists and turns in a “for and against” debate. Businesses have a shrewder sense of risk in new technology investments and will demand any greater reliance on at low code comes with controlled risks and greater visibility of what is being built and changed. However, it is well known that traditional RPA comes with increasing burdens. There is a need for greater levels of support through a constant need to repair bots, as internal and external legacy systems are being replaced and constantly updated, often without warning to downstream bots.

Often going unsaid, are the recent distractions created by RPA vendors themselves claiming to be “low code”. Whilst most visual tools could claim low code credentials since they don’t require code to make changes, it is a far cry from how enterprises are now betting their future on low code platforms for true digitization of their processes and applications.

There will be a shaking out of sub-par low code tools and a subsequent focus on using enterprise-grade low code tools that deliver those benefits of agile development but within a governance framework. Associated risks are already proving to be minimized in low code as a result and because the IT team are in a better position to oversee how, unlike with RPA, those tools are used to streamline and automate processes with a long-term future.

RPA a temporary stop gap

One of the maxims for how the IT team uses the technology is they should not make everything uniformly low code too fast. Using APIs, organizations can benefit from a hybrid low and high code model that might embed low-code capabilities inside custom-built applications. For example, this could be low-code customer onboarding applications and servicing journeys integrated inside websites developed with traditional tools and approaches. Indeed, RPA bots can be considered as a temporary stop gap to provide this integration whilst more robust techniques are generated because of more and more reusable low code assets.

Unlike RPA, where scores of bots automating legacy processes arrive in isolation through lack of governance and often with no overarching end game of transformation, low code platforms are just the opposite. Of course, even with low code, businesses, especially those in a rush, could still produce too many single-purpose apps with no vision for reuse, where each app is built on its own and then essentially thrown away. With low code though, we see more and more organizations recognizing the need for adopting a serious focus on building reusable assets that build and grow with each other and have a combined impact on innovation. This is best achieved through an emphasis on implementing low code platforms that are appropriately governed, collaborative, and follow a road map to creating an application development culture built on low code.

Another shift away from old fashioned RPA has been the arrival of low code with a brain or a strong infusion of AI within it. RPA can call upon AI, but it’s too often applied as a risky afterthought to older legacy processes that are just not ready for the scalability, security nor robustness necessary to really benefit from it versus being combined with a digital first process.

RPA has helped so many organisations achieve decent optimisations for human labour on older legacy processes. However, the limitations of managing so many standalone RPA bots have become increasingly evident in recent years. Indeed, the whole raison d’etre of how RPA bots save time and energy is undermined by how much they add to the support burden for IT teams. The restricted value of RPA means that to be justified many new plans for RPA rollouts must first come with an oven-ready plan for retiring the bots when their value is exhausted in the very near future.

So why should RPA still matter? The answer is how it is becoming an element within a larger low code technology movement where organizations can progress their hyper-automation plans in a more strategic and joined up manner. RPA is a good gap-filler but not the complete story of how digital transformation plans are being realized, grown and improved today.

We've featured the best laptop for programming.

Francis Carden

Francis Carden is VP of Intelligent Automation and Robotics at Pegasystems.