IBM RPA architecture

Learn about the IBM Robotic Process Automation architecture for all available offerings.

IBM RPA's architecture is centered around a division of client and server responsibilities. You manage administration features on the server-side. The server-side also has the implementation of specific services, like OCR and antivirus. The client-side is agent-centric, allowing main communication between client and service and providing for normal (attended bots) or headless execution (unattended bots). Business users are given a simple robot, or bot, execution application (launcher) to run bots on-demand. The integration with the client operating system, Microsoft® Windows®, makes the script authoring tool, IBM RPA Studio, a native application that runs on the developer's client-side workstation or laptop.

The nature of robotic automation operation leads to the following architectural consequence: the robot runtime, sometimes thought of as a container, is implemented on the client-side. The server-side provides all of the administration and control of robotic operation, but the actual main execution of the robot is done on the client.

IBM RPA provides two main offerings:

• A SaaS solution where the server-side components are managed, operated, maintained, and secured by IBM
• An on-premises solution where those components are installed, operated, maintained, and secured by the customer.

In both cases, the client-side components are installed, operated, maintained, and secured by the customer.

A logical view of components can be seen in the following diagram:

The diagram depicts the client-side components and features, the interactions with exposed APIs from the API server, and the server-side components.

For a detailed technical network view of the architecture, see the following sections:

• IBM RPA SaaS architecture
  The IBM RPA SaaS architecture.

• IBM RPA on premises architecture
  The IBM RPA on premises architecture.