Robotic Process Automation (RPA): Where & When – Part 2

By 13 september 2017Nieuws, Voorpagina

In this, the second of a three-part series on Robotic Process Automation (RPA), author, engineer and futurist Christopher Surdak discusses the ‘where and the when’ of RPA as a business imperative

As companies explore the capabilities of Robotic Process Automation (RPA) it’s logical to ask where best to apply this technology for maximum effect.  There is no shortage of return on investment (ROI) calculators in the world of process automation, and it would seem logical to use these traditional means to decide if RPA is worth the investment.  However, the purpose of RPA is different from traditional IT solutions, and the impact that RPA may have on organizations will often fall outside of traditional metrics. Indeed, this is what makes RPA valuable.

In part two of a three-part series on RPA, Christopher Surdak discusses where and when RPA should be used to maximize its tangible and intangible benefits to your organization.

Dirty, Dull and Dangerous

The term ‘Robotic Process Automation’ implies that this technology is something different from traditional process automation.  The distinction may seem subtle, but its impact can be profound.  With RPA, organizations are beginning to automate functions and process steps that were previously off limits to automation.  Despite decades of automation, many organizations have maintained people in certain process steps or roles in the belief that their participation was required, even mandated.  This perception isn’t exclusive to the business world.  In fact, it delayed the widespread adoption of ‘robots’ in the military for almost a century.

Examples of ‘robots’ in the military go as far back as World War I, where combatants on both sides tried to develop autonomous, flying bombs.  Such ‘fire and forget’ weapons reappeared in World War II, and evolved into the cruise missiles of the Cold War.  Reusable, unmanned aircraft evolved into reconnaissance ‘drones’ in the Vietnam War, but these were little more than recoverable missiles.

The use of drones by the military grew dramatically in the late 1990’s, as the technology improved. Again, drones were initially used only for reconnaissance, in missions classified as Dirty, Dull and Dangerous (D3). Many of these tasks could be performed more cheaply by people, but their hazardous or boring nature justified the higher costs of drones. When investing in drones, intangibles made the difference in the equation.  What is the cost to a country if one of their pilots is captured after being shot down? Only after the military considered the all-in cost of people performing D3 tasks, would drones prove to be a better option.

Unforeseen Benefits

Over time, drones have both fallen in cost and increased in capabilities and their use has greatly expanded.  Today, drones are used in ways that were highly controversial only a short time ago.  In the late 1990’s military and political leaders were deeply divided over the idea of arming drones.  The notion of armed robots going out of control and launching missiles at innocent bystanders was unacceptable to generals, politicians and citizens alike. Again, armed drones had to wait until entrepreneurs ignored conventional wisdom and fear, loaded missiles on a Predator drone, and actually demonstrated the effectiveness of this combination to the powers-that-be.

Once the concept had been proven and fears had been quieted, the use of armed drones exploded; literally.  By the time the United States invaded Iraq, armed drones weren’t just acceptable, they had become indispensable. What began as an expensive toy, derided by pilots and scoffed at by senior officers, became an irreplaceable tool of national policy.  Armed drones not only protected existing investments in planes and pilots, they enhanced their effectiveness.  Armed drones provided new capabilities, new responses and new outcomes that were unforeseen.  These returns on investment were not considered prior to use of armed drones, because they could only be determined through actual use.

The lesson of armed drones in the military is directly relevant to any business considering the use of RPA.  Many companies exploring RPA are having difficulty in establishing its ROI.  Traditional measures such as headcount reduction, improved quality or paperwork reduction often don’t justify the cost of implementing a robot. Unpredictability is a defining characteristic of innovation. As with armed drones, businesses must look to currently-intangible benefits, and potential, future capabilities, to truly value RPA as a disruptive technology.


Repetitive, Redundant and Risky

The digital workforce being deployed through RPA isn’t going to war, but it is fighting its own sort of battle.  Determining the best place to start using RPA is critical to finding early success, and to justifying further investments.  While the military has its ‘Dirty, Dull and Dangerous’ mantra, in RPA the guidance should be ‘Repetitive, Redundant and Risky.’  In looking for the best places in your business to pilot robots, use R3 rather than D3.

Any task that is performed over and over again by one or more people, with little variation, will likely be performed far better by a robot. Indeed, robots can be counted on to perform repetitive tasks much more reliably than people, as demonstrated by robots used in precision manufacturing.  If you need the same result from the same inputs, with near-perfect repeatability, RPA is your answer. The fact that robots can perform these tasks around the clock without tiring further increases their value.

The second category of tasks that are ideal for robots are redundant steps in business processes.  Despite decades of business process optimization, most processes are still filled with redundant, non-value-added steps. Where your processes require a person to review, approve, check, audit, supervise or confirm something, you have a redundancy that is crying out for robotics.

Many of these redundancies originated in our paper-based business processes, and have survived our attempts at automation. Humans make mistakes, and these redundancies were designed into these processes to catch and correct such errors. Such ‘Maker/Checker’ processes survive because of habit, expectation, regulation or fear.  Each of these factors are fairly sticky, which explains why process redundancies survive to this day.

This leads to the interesting question: Who should be the ‘maker’ and who should be the ‘checker’ in such processes? If robots can apply rules perfectly, what is the value of having a human check their work? Conversely, would it make sense to have a human be a ‘maker’ and have their work checked by a robot? Might the best ‘checker’ of a robot’s work be another robot, and would this meet the requirements of rules or regulations?  These are questions that will likely cause your organization no small amount of discomfort; just like putting missiles on drones disturbed military and political decision-makers.  Again, innovation is characterized by discomfort and uncertainty.  If your organization isn’t struggling with such doubts, you’re not innovating.

The key message with redundancy and RPA is to automate both horizontally and vertically. While there is value in replacing your ‘doers’, there may be far more to be gained in replacing your ‘watchers’, too.

Finally, the third category where RPA may shine is in managing risk, particularly regulatory risk. This may surprise many business people who are convinced that regulatory oversight demands that humans oversee their processes.  But, this may simply be a matter of interpretation.  As mentioned above, the point of many of our redundant process steps is to assure that a collection of rules has been followed or a set of conditions have been met.  Who better to apply these rules or ensure such criteria are met than an unemotional, unbiased, inexhaustible, unforgiving block of software code? One could argue that such reviews or audits are best left to robots, rather than frail, fallible, forgiving or forgetful humans.

For many businesses, process redundancies are regulatory artifacts, and they’re notoriously-expensive to maintain. Typically, such oversight is required by regulations that were crafted decades ago, if not centuries. Compliance departments and risk managers are often very skeptical of change.  After all, they are paid to be paranoid.  But the consistency with which robots perform tasks can greatly simplify our compliance obligations, generating substantial reductions in operational risk.

Not so long ago, many people argued that armed drones were illegal.  Indeed, these arguments have moved from the military to the civilian world, as a quick search of ‘civilian armed drones’ on YouTube will demonstrate.  While there may be compelling arguments against the use of robots in regulatory oversight, what may prove more compelling is their ability to generate consistent, predictable and reliable outcomes; which is the goal of any well-crafted regulation.


Adapt or Die

As history repeatedly demonstrates, there is no question of ‘if’ innovation will occur, only how quickly and how comprehensively. As in manufacturing, robots will infiltrate traditionally-human roles in services industries; this is inevitable. Some of these changes will be scary, some may even be called illegal or immoral.  But, as with armed drones, these arguments will likely fade as the benefits of innovation outweigh our comfort with tradition, or discomfort with change. It is far more likely that future regulations will tell us how to use robots, rather than telling us not to use them.



Author: Christopher Surdak