RPA Reality Check
Decoding the future of robotic process automation (RPA)
Attended Automation
An accelerated move towards attended automation is one of the trends we see unfolding in the future. Organizations around the world understand that they need to empower their employees via attended automation to automate more processes and maximize the returns on their existing RPA investments.Attended automation, sometimes called desktop automation, extends the functionality of bots even further. An attended automation robot—essentially a digital personal assistant that lives on the desktop—guides employees to deliver a better customer experience together with automating repetitive desktop tasks faster and more accurately.Attended bots take care of mundane tasks like populating forms for the employee and can also provide next best-action and compliance guidance as the customer service representative helps a customer resolve an account query, for example. This provides a bridge between tasks in the back-office and front-office, allowing more complex processes to be automated end-to-end.Consider a use case in the healthcare industry where a health plan administrator could use a mixture of server-based and attended bots to handle administrative functions like claims adjudication, authorizations and coordination of provider services. The back-office bots could get on with updating the CRM application based on details gathered by the call center agents. Attended automations, meanwhile, could help employees with tasks like switching between multiple applications, filling data in multiple places and guiding the employee in real-time to remain compliant and maximize upsell opportunities. This, in turn, frees the employee up to deal with complex customer requests more effectively or to focus on customer service, rather than on data entry.The future of RPA is smarter automations
The next step, which some organizations are already taking, is to blend attended automations and RPA with artificial intelligence (AI) and machine learning tools. Variously known as intelligent automation, intelligent process automation and cognitive RPA, this class of solutions enables enterprises to automate more complex, less rule-based tasks.Cognitive automations are able to handle exceptions and orchestrate decisions spanning entire processes, compared to RPA, which is about executing repeatable tasks with the highest level of efficiency. Whereas traditional RPA automates processes based on data in structured databases, intelligent automation can also work with unstructured data sources, including scanned documents, emails, letters and voice recordings.The technology is ‘cognitive’ in the sense that it mimics human cognitive processes like:- Learning—acquiring information and contextual rules for using the information
- Reasoning—using context and rules to reach conclusions
- Self-correction—learning from experience
Handling more difficult and complex tasks
These features of cognitive automation mean that it can handle more complex jobs without human intervention. For example, such a solution can, through machine learning, learn how to automatically address an e-mail query about a credit card dispute.Here are some examples of what the technology can already do, where an RPA platform would integrate with the below cognitive tools:- Optical character recognition (OCR) and image identification
- Extracting intent and entities
- Text analytics
- Sentiment analysis
- Categorization
- Classification
- Change format detection
- Handwriting identification
- Voice recognition
- Cognitive automation can transform unstructured data into structured data via natural language processing. This, in turn, allows many processes to be automated via RPA.
- It can handle more and more complex process paths and input types, without needing human input.
- It can learn, expand capabilities, and continually improve certain aspects of its functionality on its own.