Robotic Process Automation (RPA) has received a lot of attention in enterprise technology organizations of late. And for good reason. When thoughtfully implemented, RPA can provide meaningful opportunities for cost savings and operational efficiency. However, there are some common misperceptions to clarify and pitfalls to avoid to be successful with RPA.

​

It is important to understand that RPA is, fundamentally, a programmable tool suite configured to execute business processes by leveraging the existing user interfaces. RPA tools engage the same UIs that human operators engage. A RPA bot will populate text fields, click buttons, generate spreadsheets, and send email attachments in much the same way as a human. The primary benefit of RPA is that it enables an organization to automate processes across multiple systems without investing in expensive, multi-year back-end system integrations.

​

It is also important to recognize that RPA is not Artificial Intelligence or Machine Learning as we’ve come to understand those terms. RPA bots do not “learn” and are not capable of making decisions. Though highly configurable, RPA bots do not improve over time without being reprogrammed. That said, a combination of advanced AI and RPA can have powerful results.

​

Because RPA bots engage with the same UIs as human operators, the speed with which a bot can execute a process is dependent on the speed and responsiveness of the UIs being engaged. The fewer the steps between input and desired output, the more effective the bot. Thus, any successful RPA implementation hinges entirely on the consistency and efficiency of the business process being executed. In this regard, RPA should be thought of as a technology supplement to a robust process improvement framework.

​

A successful RPA deployment should begin with an in-depth Design for Six Sigma (DFSS) analysis. In concert with an experienced DFSS practitioner an experienced RPA developer will perform a keystroke-level analysis of a human operator’s process -- along the way identifying inefficiencies and communicating the capabilities and limitations of the tool.

​

When well configured, RPA bots can execute simple tasks much faster than human operators, with greater accuracy, and with near 100% availability. But like any other tool, they must be set up for success.