By using Robotic Process Automation (RPA), we are able to create a digital workforce. But how autonomous will this workforce be? And how can we manage such a workforce?

<<< Start >>>


<<< End >>>

"For our rapidly growing organization, we are currently looking for claim handlers. Your work will be to ensure that insurance claims are handled efficiently and that payment for valid claims is made to their policyholders. You need to decide on the extent and validity of a claim, checking for any potential fraudulent activity and further process the claim. You will work 24 hours a day for 7 days a week and you will cost us 75% less than the workforce you will be replacing."

RPA Is Shaping the Workforce of the Future

Well, this is a job offer no robot can refuse. Robotic process automation (RPA) is quickly shaping the workforce of the future. Gartner1 predicts that current RPA vendors will add more process or task-specific capabilities to their software, while in the meantime vendors in the process domain will add RPA-like functionality. This suggests that in a couple of years most of our human administrative workforce would be replaced by robots.

The question is, however: Are robots suitable for every job? Can we simply replace every human worker by a robot or will the future offer us a workforce that exists of humans and robots working together?

In order to effectively build up our new workforce, we need to know the key strengths and weaknesses of robots in different roles and with various responsibilities. Based on that we can decide whether or not a robot is the perfect candidate for a job.

Let’s take a look at what I consider to be the top 5 of pitfalls in using robotics in business processes.

1. There Is a Robot for Every Job

It is part of my job to visit clients and perform an assessment of processes, to determine whether or not these processes are a candidate for automation. Could I be replaced by a robot?

In case of a straight-forward situation: Yes. I could even program the robot myself, using an expression with a few 'if-then statements' and a decision based on a simple calculation. The result: a robot that helps you robotize your processes! Faceless, fearless and without any emotion.

But imagine now the impact of taking humans out of this process for both employees and their managers. It would almost be like going to a doctor for a test and getting the diagnoses on a computer screen. Robots cannot reason and cannot look beyond the directly calculated impact of their decision.

A robot is often not the best fit for jobs where interaction is desired, reasoning is required that goes beyond simple if-then statements and a holistic view on a situation is necessary. There are things a human is better capable of in certain situations. It is therefore simply a fact that in any business process a human factor will remain. Robotics could be a mean to reach a goal, not the goal itself.

2. One Robot That Does It All

We can make a robot as complex and extensive as we want. One robotic process could take over the administrative work of an entire department, encapsulating the complete process with all its exceptions in one flow. Think of a robotic process that picks up tasks from a system one by one, determining in the first step what the task is and then choosing a separate flow per task with branches for every exception. The impact of changes in either a process or a system would be huge; the robot would stop functioning and no single task could be performed anymore. In addition, such a robot requires logic that exceeds the typical simple if-then expressions, making it even harder to fix a problem when it arises.

By designing small tasks, that can be scheduled and assigned to a robot, we can make the life of the maintenance team easier and provide organizations with a robotic workforce that is more stable.

<<< Start >>>

"RPA has never been designed to fully replace humans"

<<< End >>>

3. Robots That Reason and Take Cognitive Decisions

With the demonstration of technologies like Watson and IPsoft Amelia in mind, we kind of expect that the current generation of robotic tools is capable of reasoning and handling complex decisions independently. But what we shouldn’t forget is that RPA tools are designed to mimic human actions in repetitive processes. Complex decisions will result in long and complicated series of expressions in a flow, leading to multiple flow options and potential exceptions. Maintenance and troubleshooting will be a nightmare once you incorporate these kinds of complex decisions into an RPA task.

RPA has never been designed to fully replace humans, they are designed to take the robot out of the human and leave the more interesting and interactive work for their living and breathing colleagues. Of course, we can expect a growth path from the current RPA tools towards more cognitive solutions. The Accenture Cognitive Robotics solution is a good example of a solution in which robotic and cognitive technologies are combined.

4. Robots Can Process High Volumes in Real-Time

A robotic process is typically designed to handle a batch of similar activities within a pre-defined period of time. Tasks are scheduled and are performed by using the graphical user interface of existing systems. By using robotics, organizations are capable of increasing their processing speed up to the limits of what the underlying systems can handle. In fact, the processing speed of the robot is often determined by the loading times of the interfaces and the processing power of the systems underneath it.

If you want a solution that is up to 4 or 5 times as fast in performing a task as a human workforce, look no further. But if you want to process high volume tasks in real-time, a service-based solution might be a better option.

5. Robots Don’t Need to Be Managed

Robotizing one robot and bringing it to production is one thing, but what if you have to robotize hundreds of processes? Who would be managing these robots? By introducing robotics in business processes, we often introduce new types of repetitive work: monitoring RPA dashboards and taking action when an error or exception arises. What we should realize is that we need to think in solutions that are capable of handling a hybrid workforce consisting of both humans and robots. A robot – so to speak – that manages the work, orchestrates processes, assigns tasks to robots and human experts and keeps an eye on the timelines. Just like Obvion CEO Ronald Touwslager said, it's about jointly defining the way forward. "I really believe in the power of ‘together’, also with respect to our partners, like Accenture with whom we designed our robotics model,” Touwslager believes. "Accenture was open to different collaboration models and was prepared and able to stick to transition pace we sugegsted.”

<<< Start >>>

“I really believe in the power of ‘together’, also with respect to our partners, like Accenture with whom we designed our robotics model.”

<<< End >>>

RPA Orchestration

This is basically the idea behind the Accenture Intelligent Process Automation platform in which we combine RPA with a cognitive reasoning engine and dynamic process management.

This platform offers an eco-system that combines the proven effectivity of flow-driven robotic solutions with a goal-driven orchestration layer and artificial intelligence. This orchestration layer can handle complex decisions, classifications and calculations and trigger both manual and robotic processes. The addition of artificial intelligence makes it possible to create virtual assistants that interact with customers and enables cognitive – self-learning – robotic systems. The result is a flexible and future-proof solution that enables organizations to start small – with the easy to automate processes – and expand gradually towards the more complex processes.

Designed by humans, made for humans and robots, automating business processes from beginning to end.


  1. Gartner: Market Guide for Robotic Process Automation Software - Cathy Tornbohm (7th of November 2016)
Subscription Center
Subscribe to Accenture Insights Subscribe to Accenture Insights