Unlike Macros, Robotic Process Automation tools can be used to perform end-to-end process automation. The complete automation code is called a (ro)bot. It works at the UI level also called as surface automation.
In RPA, bots are universal application orchestrators – any application that can be used by a person can be used by a modern robot, whether mainframe, legacy, bespoke application, web service enabled or even a closed 3rd party API hosted service.
Bots collect procedural knowledge which over time build into a shared library that can re-used by any other robot or device (in the same way objects are built in traditional Software Engineering).
Applications are “read” by the bot, either through dedicated APIs where they exist, through the OS prior to application display, or through the screen in the context of the native application. In last case, the modern bot “reads” an application screen in context and in the same way a user does. As part of the robot training, it is shown how to read the application’s display much like a user is shown.
|Learns and enhances itself from the repetitive process||Does not learns from the repetitive process|
|Can act Autonomously||Cannot act autonomously|
|Responds to external stimuli and reprograms itself||Does not respond to external stimuli|
|Highly secured automation||Security is not a high priority here|