According to Gartner, 100% of Fortune 500 companies have automated processes. Out of those, about three quarters are actively running automation projects of some sort. RPA’s growth is taking over portions of legacy automation by converting to RPA. This results in highly increased efficiency.
Automation is not new, we have been moving in this direction for the last half century, if not longer. In its initial automation waves, manual labor was automated with robotic arms. The most well known example of this is the car-industry. Fast-forward to the last few decades and within the information-worker’s space, we see an important increase in conversions from automation through batch processes into processes on demand and data transfers, data conversions, and data transformations. Finally we see those convert into smart data flows which finally evolved into the current RPA wave we have today.
During the 1990’s data transformations were left for DBAs and specialized technical/business analysts. These professionals would make sure disparate systems would be able to communicate with each other. Building data transformations would require getting the proper analysts to agree what one was going to export for the other to import.
There would be complex processes to extract data from one database and import to another. You need to program, QA test and implement these processes. Data exchanges would not be a one-sided conversation; it very much required two sides (one for each system) to work in a coordinated fashion with a third party that would produce the solution (sometimes a system onto itself) that would execute the actual process.
Enter RPA tools such as UI Path, with these modern automation systems, the process of automating only requires of a UI. Automatic transformations, conversions and the system’s ability to conditionally execute one or another part of the workflow allow for a much more flexible process. RPA processes are more nimble and able to protect themselves from failure by modern error handling.
In the past, having a “single system for running all automation in the Enterprise” would be frown upon. Building for all possible systems that one would want to automate was considered very impractical. Even more, be able to capture, screen-scrape or extract values would be near-impossible or too cost-prohibitive. Today’s market offers several products that while focused on the same goal achieve these goals in different ways. You don’t need to use different RPA solutions to automate different systems, nor you need complexity to exchange data between them. As long as you know how to use the UI of a system, you can either extract or feed data to it.
Most RPA processes benefit from the use of triggers that start workflows on certain conditions; for example, when an email with subject “Invoice from XYZ” arrives. This is similar to using inbox rules, but the interesting part is what you can do with the data included in the email subject, body, and attachments.
When looking into a UI to capture a value, RPA workflows can even wait until a particular element appears (or disappears) from screen allowing the process to pause until that event triggers it to continue.
Finally, there are plenty of SaaS (Software as a Service) tools that run on a browser. We store this data in the cloud and sometimes we have a limited number of extraction options. Data extractions are made, purposefully or not, complicated to extract to other systems. RPA tools can extract data from a browser session as easily as it can extract from an application running on your desktop bringing back control to the Customer over its own data.
Multiple events could trigger RPA workflows. In the previous invoice example, while some of your vendors could email you a PDF with an invoice, others may send you an application notification with a link to download a report on a spreadsheet. RPA can handle most scenarios. Chances are, if the application to review the data sent to the information worker is usable by a person, then you can automate it with an RPA tool. In absolute worst-case scenarios, the RPA tools can screen-scrape with OCR (optical character recognition). The tool then imports these OCR results into the system for later use. Just imagine yourself reading the screen, and then typing the text from into the next system. This process works just the same way, but it does everything for you.
RPA tools have evolved, they are easy to use, have incredible flexibility, and their implementation normally results in an immediate increase in efficiency. It took a long time to get here, but it has landed as a useful way to reduce times and increase efficiency without sacrificing quality.