Artificial Intelligence, Automation, Robotics are enabling the 4th Industrial Revolution. Robotic Process Automation (RPA), a catalyst for AI adoption, plays a central role in this evolution, by empowering organizations to take their Digital Transformation to the next level.
Robotic Process Automation is one of the latest components added to the Process Management landscape. Even though its market took off ~5 years ago and grew exponentially, I think that it is still not clear today for many people what RPA represents.
For starters, RPA has nothing to do with the robotics industry: there are no physical robots that interact with real-world objects. What we actually have are digital agents / applications that execute a series of operations, both in the background and via the user interface, in order to perform a predefined task, process.
UI actions are very common in RPA workflows, leading people to think that Robotic Process Automation = UI Automation .. and that is not the case: digital processes contain a lot of steps implemented with background running actions, not on top of the UI; for example:
- sending an email
- making a request to a web service (I used the Microsoft Cognitive Services Face API in my demo workflow)
Today’s RPA platforms make heavy use of Artificial Intelligence / Machine Learning capabilities for both UI and background actions:
- identification of user interface elements for UI Automation
- data extraction from images (bills, contracts, ..)
- text processing (e.g. interpret and process service desk requests)
Robotic Process Automation significantly reduces the entry barriers for Artificial Intelligence adoption by making it straightforward even for non-technical users to integrate machine learning components into their process automations. RPA is integrated in organizations mainly because it delivers hyper-efficiency and very fast ROI, but it can also act as a main pillar for an AI driven Digital Transformation Strategy.
I have uploaded my presentation on slideshare and it is also embedded below: