Robotic process automation (RPA) is a technology that makes it easy to build, deploy, and manage software robots that emulate human actions by interacting with digital systems and human users. RPA is used to automate operations and streamline processes. Companies often deploy a hybrid workforce that includes both humans and robots to collaborate on processes that can be semi-automated but still require a human at key decision points. RPA is used to automate supply chain processes, including data entry, maintenance and after-sales service. The business case makes most sense for high volume, rote tasks that are highly standardized. For example, telecommunications companies use RPA to configure new services and associated billing systems for new accounts.

There is a eight-step process for planning, implementing, and improving RPA systems. 

1. Prioritize processes to automate: Not all business processes are suitable for RPA. Define the purpose, business context, and impact on business operations. Then prioritize processes based on metrics like frequency, complexity and ROI. 

2. Perform feasibility assessment: Assess the feasibility of RPA for each process. Some processes can be fully automated, others can be partially automated. 

3. Restructure processes: Based on the feasibility assessment report, identify processes with a high expected ROI that are not yet sufficiently standardized. Optimize and restructure the process before implementing RPA.

4. Gather user stories: A user story describes the features of an application to be built from an end-user perspective. Based on user stories, develop a process definition document with RPA workflows.

5. Develop the RPA software: Developers create automated scripts and program code using tools like UiPath and Blue Prism. 

6. Test RPA processes: Perform tests to assess performance in all possible scenarios and send performance issues and bugs to the development team.

7. Confirm and deploy: Once all issues are corrected by the development team, confirm the results are sufficient and deploy the complete RPA solution.

8. Improve continuously: Many RPA systems are based on machine learning rather than hard-coded processes. These systems can be improved as additional data is ingested.