For many companies considering ways to automate business processes for more productivity, chances are Robotic Process Automation (RPA) is one of the first technologies they’ll look at.

Some of the quickest and easiest opportunities to automate a business – the “low-hanging fruit” – will come from RPA. It’s one of the smartest technology evolutions in recent years, offering a way to reduce costs, enhance productivity, and even increase efficiency and accuracy.

Yet for many businesses, there is a good deal of confusion about what RPA can actually achieve and what it’s limitations are. When considering implementing RPA, it’s best to have clear expectations so that you can best determine how to leverage it for strategic advantage.

The best way to understand the opportunities around RPA is to imagine a human employee carrying out dozens of prosaic, repetitive tasks in their daily routine. Over time, these tasks become dull and tiring, and human workers become less productive or make mistakes.

RPA executes these tasks more quickly and accurately, without tiring. The human worker is unleashed to focus on tasks requiring more critical thinking, creativity, and higher-level engagement with customers.

A great example of a process that’s a prime target for RPA-based automation is an invoice verification process. This process could involve a human user downloading invoices from three different vendor portals, uploading them into an ERP system and verifying the values with the corresponding purchase orders, and eventually approving or rejecting them on the system. This process can be very easily incorporated in an RPA software with some amount of scripting.

RPA is a way of applying technology to configure a robot or computer to use and interpret existing applications for processing a transaction, manipulate data or trigger responses and communication with other digital systems.

To simplify, RPA is a “record and play back” technique where the robot (which actually is software installed on a machine), is taught how software is used to carry out some tasks and the robot does it recurrently, making it a simple yet effective automation method.

Traditionally, automating that same invoice verification process would involve integration with three different vendor systems, which must be integrated with the ERP system with some verification logic. While the outcome of both processes would be the same, the traditional route of automation comes with the added tasks of software development, working with the IT team to open ports on firewalls, having servers dedicated to run the software, and some maintenance effort.

RPA is a much simpler process. The software can automate numerous other processes as well, as it can run 24 x 7, and most likely the invoice verification only needs to be processed for a few hours every week.

Roadblocks for RPA

While RPA provides a very quick, non-intrusive means of automation, it has its limitations. To maximize RPA’s usefulness, the processes being automated need to be based on well-defined standard operating procedures (SOPs) where the rules for decision-making have very clear parameters. In any RPA-based automation strategy, it is important to split the larger process into “automatable” and “human intervention” areas and update the process to enable maximum automation. However, it may not be practical to completely automate every process This is because some processes are fairly procedural, while some may require applying a human worker’s knowledge and making judgments. In cases where vast amounts of knowledge cannot be effectively codified into algorithms, RPA will not be able to automate certain processes, which will need to be routed to a human user. For areas where human judgment needs to be applied, RPA is not effective.

The automatable part of the process that is driven by RPA thereby becomes completely rules and algorithm-driven, and it is hard to make them “smart” or “intelligent.” In certain cases, this can shrink the automatable part of the process to a level where it becomes non-feasible to apply RPA as the returns may not justify the RPA investments.

An example of this would be automating the process of routing problem tickets to the appropriate groups. If the routing is based on ticket category and certain fields of the issued ticket, RPA can certainly solve the automation problem. On the other hand, if it involves reading the content of the ticket description and applying human knowledge to make a decision on which group should handle that ticket, then suddenly automate the process looks non-feasible by RPA alone.

How Machine Learning (ML) Can Help

Machine Learning (ML) is an emerging technology that is already maturing to a level where it can be applied to solve real-life problems. ML works on the principle of capturing a large amount of data (or knowledge) into some form of mathematical model. The model can be utilized to apply knowledge for solving problems.

For example, if we develop an ML model with one year’s worth of ticket descriptions and its corresponding routing information, it can be used to predict the routing information for a new ticket. ML has the ability to learn based on prior actions and apply that knowledge to make a decision on current actions.

ML can be applied to a multitude of problems where there’s access to large volumes of historical data which can predict or make decisions on certain areas. It is unlike developing an algorithm; it is based on building a knowledge base. If we take the ticket routing problem, it will be extremely hard to develop an algorithm taking into account every ticket that has been raised in the past one year. However, the ML approach can address such scenarios very effectively.

The Powerful Combination

Combining ML with RPA can help to overcome RPA’s limitation, as it can build a knowledge base from historical data, and use it for decision making and prediction.

If we take the same example of a process of ticket routing, it now appears completely feasible, with the RPA process taking the ticket description and making a call to the ML prediction service to fetch the routing information to continue the automation process. It is like introducing a smart agent into the otherwise procedural RPA process. It is also feasible to keep enriching the ML model by re-training it periodically and keeping it powered with the latest knowledge.

Robotic Process Automation is an exciting technology on its own and coupled with Machine Learning, the possibilities to drive savings and efficiencies are endless. While the combined technologies make a powerful automation solution, there is some work to be done in making it accessible to an end user. However, with the right expertise, enterprises can realize truly intelligent automation solutions that will garner the results they have been looking for.