This paired couplet of guest posts for the Computer Weekly Developer Network is written by PK’s Ram Sathia in his position as VP of intelligent automation and Ajmeer Ali Liyakkathali in his position as PK’s associate director of intelligent automation.
PK describes itself as the ‘experience engineering’ company — its software works in areas related to marketing touchpoints that enable brand connections like lifestyle apps, 5G and IoT transformations, personalised e-commerce and connected health care platforms.
Sathia & Liyakkathali write as follows…
Reports suggest that more than 33% of companies today have piloted or implemented a Robotic Process Automation (RPA) initiative. But not all projects are successful; in fact, some studies have found that RPA projects miss deadlines more than 60% of the time and have failure rates of 30%-50%.
This is because achieving ROI with bots depends on not only developing and implementing them but also operationalising them. To operationalise bots, you must develop them in a manner that makes the technology effective, scalable and easy to use for manual tasks.
Bot operationalisation can be challenging, but developers can follow some key best practices to put bots into action in everyday work.
Infrastructure & environment
There are a number of potential failure points for bots in production, such as a lack of consistency in OS, browsers, RPA tools, license server configurations, security patches and IAM rules. These are often difficult for RPA developers to debug. Leveraging containers for automated bootstrapping and provisioning of RPA infrastructure builds a solid foundation for successful bot development and operationalisation.
Developers can also use containerisation technology at enterprise-grade to reduce infrastructure support and downtime for RPA initiatives.
Focus on business objectives
In the development phase, it’s easy to fall short of business objectives. Common missteps include failing to create a standard framework, using incorrect or unscalable component design, or not fully understanding how a bot must function in a process.
Often these problems could be avoided with the right framework and best practices to ensure that bot projects meet organisational goals.
It’s also important to develop a framework for scalability.
Automate your business’ automation of manual processes by building a framework of commonly used libraries that can be selectively changed by additional user-written code. This makes it easy to create an application-specific bot and achieve higher ROI.
In any RPA process, there are many common actions/applications that developers can automate and leverage across multiple bots and business functions. One example is email automation, which includes reading email, downloading attachments, sending email, replying to email, moving email, and deleting email.
Simplify component design
Often RPA is developed with an excessive number of commands in component design, which leads to overly complex bots that are difficult to implement. When designing bot components, always stay as simple as possible and include only what matters to the components. This often means building a combination of components that fulfill functional needs and are large as well as small single components that increase development costs but enable flexibility with fewer commands. Taking this approach provides a middle ground where costs are reasonable and bots are still easy to use.
Use a layered approach to component structure is also important. We recommend structuring components in layers of blocks such as framework, applications, data, microservices and application screens. Keep business logic outside of these basic building blocks. This allows components to be reused across bot initiatives, which helps reduce time spent on code development and overhead costs, and increases cross-compatibility between bots.
Have an RPA architect review bots for functionality and best practices. This process ensures and enforces the use of best practices while helping to avoid any gaps in development.
Adopting these recommendations empowers organizations to quickly and reliably develop and deploy bots, maximising the ROI achieved from automation. In our next piece, we’ll examine how to build bots in a way that allows tracing and recovery from failures.