In 2005, the robotics company we started that was acquired by Pega, pioneered a unique approach to automating the desktop user interface. Our approach utilized a well- documented “secret” technique of “memory injection and hooking” that enabled humans and software robots to work side by side on the same desktop, without stepping on each other’s toes.
This became more commonly known to our customers as “deep robotics” and it was significantly faster and more robust than other common approaches to automating desktop UI at the time. Since OpenSpan’s acquisition by Pega, this technique has never been replicated or surpassed by any Robotic Process Automation (RPA) vendor.
With attended RPA built with deep robotics, customers with tens of thousands of users are able to deploy an equal number of bots, with each bot deployed and able to update itself automatically. This gives enterprises the ability to reduce call times and transaction times, speed up employee training, and eliminate errors and rework while getting new users up to speed in a fraction of the time it used to take.
I give you this background to explain how Pega RPA was designed to be uniquely different than other options in the market from the outset. And over the years our approach has proven successful at deploying tens of thousands of bots across many of the world’s largest enterprises. While we’ve always supported both unattended and attended RPA, our core focus has been on perfecting the hardest parts of UI automation first – namely, how to deploy bots in real time, on the same machine as the user, without fragility or brittleness.
Why then, is attended RPA such a powerful necessity? The story begins with the Microsoft Windows operating system, which opened the door to create what became a common industry problem, with its uniquely open environment. Windows encouraged application UIs to be built, but often each was built in a very different way. So we saw enterprises expand, ending up with 10 or more of these applications at the same time, on a single desktop. Then the problem became the fact that the only way to make these applications talk to one another at the desktop, was to pay a human worker to act as the expensive integrator, manually navigating their screens, keyboard, mouse, and (sometimes exorbitant use of) commands like alt-tab and copy-paste.
What this means for businesses is that thousands of users in the front and back office spend between 25 and 90% of their time on every call or transaction doing this manual work navigating between apps – and these costs can add up significantly.
Because the volume of transactions and calls executed through the desktop in an enterprise’s front and back office are often still massive, this is where attended RPA using deep robotics makes the biggest difference when compared to traditional RPA UI automation techniques. Traditional RPA and older techniques, such as OCR, send keys, and human in the loop unattended RPA, can cause user frustration with delays, screen hijacking, and the inability to work faster than the employee can. Deep robotics RPA, however, takes a different technical approach to deploying its bots. And that approach can be over 20X faster than a human and 10X faster than traditional RPA.
We invented deep robotics for attended RPA to address this, and we now have customers that started with us more than 15 years ago that are still finding a need for attended RPA. The bottom line is, if you have a desktop with more than one application and humans are the integration point, then this need for “RPA done right” should continue to be in your automation arsenal.
To get more information about Pega RPA, feel free to reach out to me at [email protected]. I'd be happy to share more about our approach to robotics with you.