In this, the first of a three-part series on Robotic Process Automation (RPA), author, engineer and futurist Christopher Surdak discusses the ‘what’ and the ‘why’ of RPA as a business imperative
For business people used to an endless parade of jargon, hype and hyperbole, the term Robotic Process Automation (RPA) may sound like yet another label worked up by yet another high-performance, cross-functional tiger team from marketing. While a degree of cynicism may be warranted, few would argue that businesses need to change their approach to automation and business optimization in order to remain competitive.
RPA is not as much a new set of technologies as it is the application of technology to a new set of problems. Historically, information technology was used for the processing, distribution and control of data. Capturing, evaluating and auditing information, before and after processing, typically fell to humans. Whether standing in line at a bank, an airport ticket counter or at a grocery store checkout lane, we still have millions of workers world-wide who spend their day staring into a computer screen and contributing to the processing of data.
With RPA, businesses are looking to replace humans with software in these low-value-added process bottlenecks, with the goal of improving quality, speed, efficiency and outcomes. Cost has been and remains a major driver of automation. However RPA promises much more than simple cost cutting. RPA also allows organizations to redeploy people into high-value added roles that require judgment, empathy and cognition; things not readily delivered through machines.
Haven’t We Been Here Before?
Process automation is nothing new; businesses have been doing this for decades. For up to half a century companies have expended a vast treasure on using IT to improve the quality and predictability of their processes. From Kaizen to outsourcing, from Lean to benchmarking, organizations have tried to make their businesses more efficient and less expensive by first automating and then commoditizing their old paper-based processes.
This provided significant gains in productivity since the early 1990’s and led to the business process outsourcing industry, worth over $100 billion in 2015. However, many companies who embraced these initiatives have found that there is little room for further improvement. Many of us have reached the point where not only have we picked all of the low-hanging fruit from our old manual processes, we’ve stripped these trees bare.
There comes a point where constant incremental improvement necessarily approaches perfection, and there’s no room left for improvement. After decades of effort, many companies have reached this point; still others have died trying. Not so long ago, Six Sigma was an operational goal. In today’s connected, app-driven, instant gratification world, achieving merely Six Sigma quality would be a sign of weakness, customer frustration and shrinking revenues. Nine Sigma is a bit closer to customers’ present expectations, and when an exception occurs, you better know about it before your customer does.
Recent advances in componentized software allows for much faster design, deployment and implementation of a digital workforce. While prior efforts at automation focused on data processing, RPA focuses on automating the steps between automated processes. Historically, these steps remained with human operators as a matter of convenience, cost, trust or habit. Tasks such as validation, inspection, data entry, review or approval were performed by people at the boundaries of automation, giving people the perception of control.
As the volume and velocity of business have increased, performing these steps has now grown beyond human scale. Rather than contributing to process outcomes, humans are rapidly become the limiting factor in further process improvement. Our data processing systems have grown so powerful that processing takes almost no time at all, it’s the steps in-between that customers notice, and dislike. Today, handing your credit card to a cashier, so they can manually enter your security code into the register, takes far longer than it takes to receive approval from the card issuer.
Why Not Reengineer?
Many of these human-centric steps seem natural targets for reengineering. But, those who have tried reengineering legacy business processes know how difficult reengineering can be. Many of our business processes are reinforced by decades of habit, expectation and regulation. These are things that do not change easily, quickly or inexpensively; as anyone frustrated by using a self-checkout machine at a store can attest. Business processes could be reengineered to remove these steps, or to directly integrate different business systems. However, reengineering is costly, messy and extremely disruptive.
What organizations need is a means of automating low-value-added human activities in business processes, without directly integrating the legacy systems that underlie these processes. Fortunately, with RPA such automation is not only possible, it’s preferable. Robots are able to perform the same steps presently performed by people, but can do so without tiring, without errors and without distractions from social media updates. When a process requires modification, one need only create an update to the RPA rules, rather than planning an employee training campaign. Businesses can implement RPAs on their own schedules, adapting to local customs, preferences or regulations as required. Finally, RPA facilitates large-scale reengineering by identifying and eliminating barriers to change one small increment at a time.
Everything Old is New Again
In some respects, RPA is nothing new. Businesses have been adding technology and automation to business operations for decades in order to reduce costs and increase efficiency. However, many process steps remained in the hands of humans, because these steps demanded either flexibility, cognition or oversight. Software has advanced to the point where these process steps can also be effectively automated, which may lead to another generation of process and productivity improvement. The key to effectively implementing RPA is to understand its benefits and its challenges. These include:
Benefits of RPA:
- Existing processes remain intact, no need to reengineer
- Existing controls, reporting, auditing, etc. remain intact
- Immediate improvements
- Deploy as quickly or as slowly as the organization chooses
- Allow local variation in processes
Challenges of RPA:
- Defining business cases with supporting Returns On Investment (ROI)
- Preventing scope creep and added complexity
- Automating inherently-poor process elements
- Managing exceptions when they occur
Organizations that deploy RPA effectively can redeploy their people into much more value-added roles. This can increase ROI, employee satisfaction and customer satisfaction. RPA should also be viewed as a necessary complement to an organization’s effort to leverage machine learning (ML) and artificial intelligence (AI) for even further gains. Orchestrating RPA and AI efforts will be addressed in later