5 Concerns about Robotic Process Automation

CXO THOUGHTS
Sonia-Clayton-2020-70-4 5 Concerns about Robotic Process Automation
Sonia Clayton, CEO, VIP-Global

For those of you who are learning about Robotic Process Automation (RPA), you must know that it is a great replacement for mundane and repetitive processes. Failure will occur and when that happens the question is… “are the processes right?”

  • The process: If the process variants and tasks are wrongly applied and programmed within RPA, the project most likely will fail. Also, keep in mind that tasks that require judgement even if they are basic will not be handled well by RPA alone, nor is a task programmed with a poor process. If the processes are broken, RPA won’t achieve its objectives, so we must pay attention to exceptions and variants.

David Brain [1], makes a clear recommendation about this topic and he says we must look for predictability, repetitive rules, and frequency of mundane tasks. Without exception, variable or changing processes or non-predictable ones will not be a good fit for RPA. Failure can happen when decision-makers conceptualized processes that may not necessarily be robotic processes, demonstrating that wrong processes can be automated. Should this be the case, all RPA tools will require more maintenance and development, or they may simply become disabled. Metrics of high maintenance and development or just failure will not look good on the ROI of the implementation.

  • ROI of RPA: Antony Edwards [2], says that measuring the financial ROI of an RPA implementation can be boiled down to a before-and-after – or old-versus-new – comparison and final results. In other words, the old means how much effort was it taking before multiplied by the salary of the people doing the work. And the new is equivalent to cost of RPA solution, plus the cost of people managing the RPA. So, if the “new” cost is less than the “old” over similar timeframes there is proof that the RPA project was worth it since it reduced and achieved the expected financial objectives.
  • Too Much Human Interaction: If a process requires human decision-making and creativity, it is probably because the environment of the automation scenario is much more dynamic than initially expected. Hence, there are more RPA tools involved that may require more human interaction which defeats the purpose of RPA because we are creating more complexity. The point of RPA is to minimize human tasks, so humans can concentrate on relevant tasks to generate better results, in less time.
  • RPA and Artificial Intelligence (AI): RPA alone is repetitive and does not learn on its own, nor does it provide responses to unexpected events. That is why Artificial Intelligence (AI) is needed. The effectiveness of RPA and AI is the actual revolutionary solution because when they are deployed together is when we enter the world of cognitive technologies.

During my last visit to the City of Prague, I interacted with an RPA bot called ‘Pepper” that was programmed to meet and greet customers. Upon greeting Pepper, we first had a brief conversation about travel. But then, I changed the subject and Pepper went silent, because what I was asking was not part of Pepper’s programmed responses which would have required analytical thinking. Pepper was not trained [for] or instructed on how to manage that situation.

  • RPA-OCM: Robotic Process Automation Organizational Change Management? If your leadership is willing to take the risk with innovative technologies, most likely your project will succeed. But if the environment is highly political in your organization, then your RPA implementation is at risk. It is a well-known fact, that your technology investments more likely will succeed if your executive champions are leading the change and transformation. Other than that, it will be very difficult to demonstrate the value of the transformation.

And who is managing the panic button? According to Harvard Business Review, most operations groups adopting RPA have promised their employees that automation would not result in layoffs.[3] Instead they explain that most workers will be redeployed to do more interesting work.

In conclusion, before implementation, we must have a good idea of Cost Vs. Benefit. Having a good understanding of RPA and how its implementation makes sense for your organization is key.  A good RPA implementation requires leadership that supports OCM, experts who can measure and report value-added metrics, who can document and implement good business processes, maintain tools, and train with AI (Machine Learning) respective and cognitive tasks. Keep in mind, that RPA will not displace workers but will enable them to focus on higher productivity and better quality.