Process Automation is not a simple replacement to the current manual processes followed. Intelligent Automation is aptly illustrated by EY in this report, “Intelligent Automation – Reshaping the future of work with robots”
- At the beginning of this transformation is the collation of the current process and clustering them into Functions.
- The next step of Standardization involves identification of differences within each process step and unifying the process based on best practices.
- Optimization involves a critical review of the end-to-end process as well as a deep review of each process step to make it lean and streamlined. This step requires individual owners to discard their prior notions and adopt a new lean principle.
- At the final stage of evolution is the application of Automation to the optimized process
Organizations often resort to shallow record-playback approaches to Automation. Automation, in itself is a Software. Automation needs the same rigor of Architecture and Design considerations as that of Software. Functional and Technical modules derived in the Optimization step above needs to be designed separately. Concepts such as Open Interfaces and Loosely Coupled Design are important. Every process step like Data Entry, Data Retrieval, Web Login, Web Scraping, HTML Parsing etc should be designed as individual building blocks controlled by XML metadata for configuration. This will allow the reuse of each module across multiple websites and data schema during custom implementations. The output of this design exercise is a collated library of Automation Modules. Each customer implementation will involve a review of the process steps and stitching together of the right modules in the necessary sequence, setup with appropriate XML parameters.
Base Tenets of Automation
- Modularity – Automation must be modular and allow for easy improvements and extension without redoing the whole project. Wherever possible, API and callback hooks should be provided in the core modules
- Extensibility – Each module must be designed with extensibility and Configurability at its core. Effort must be placed to externalize variables, data, messages and errors that will need to be modified with each implementation
- Maintainability – The single most point of failure of Automation is maintenance. Since Automation projects are usually hard-coded without attention to future change management, it keeps breaking for too often to justify its value. This is where rigorous software design principles are required
- Cleanliness and Hygiene – Much like complex software, Automation needs to be cleanly separated with code comments and Source Control Repository for maintenance and versioning
Usage of Tools
Automation uses many organic technologies like Seleniun IDE, which is a time-tested UI Automation framework. There are other similar open source tools like UIVision. Commercial Automation tools include big players like AutomationAnywhere, BluePrism and UIPath.
In this Video, Edureka has compared these three commercial tools along fifteen-point criteria
In conclusion, Process Automation is a special project that requires the best process experts, lean thinkers, and software architects to come together to build a lasting and valuable Automation platform. Automation is as complex as building the right software and organizations should champion it the same way as they do complex software projects.