RPA bots imitate human interaction across different screens and/or applications without changing the existing IT environment. Classical RPA is limited to structured and relatively simple processes where little to no interpretation is needed, but what if we link up AI in the same process? Then we’re talking about cognitive RPA or Smart Process Automation.
As the term indicates, cognitive RPA can best be regarded as a cognitive layer on top of a classic RPA process. But how is this done? By making the link with Artificial Intelligence. Thanks to AI you can also automate processes where semi-complex interpretation is required.
An example of cognitive RPA
So what are ‘processes with semi-complex interpretation’? Take the following simple example: You work in customer service and receive many emails every day with questions and complaints. The content is very diverse: from dissatisfaction with a product that works badly to a person who has forgotten their password. And as is typical for this type of communication and in particular for complaint emails: they are often drawn up unstructured.
Via RPA you can let go of natural language processing on those mails. Based on the intelligence that lies behind it, the bot automatically determines exactly which topic the question or complaint is referring to and in which language the email is written. The bot will then act – according to a script –and immediately send the question or complaint to the right person. There is no need for manual intervention and the emails get to the right person faster.
RPA as an enabler for intelligence
The example of natural language processing is only one of many options for cognitive automation. Cognitive automation imitates the way a person thinks and automates tasks that do not follow a fixed set of rules. Technologies such as pattern recognition, natural language processing (NLP) and image processing are used when interpretation of information is required, as in the example above.
In other words, a classic RPA bot copies human actions, while Artificial Intelligence simulates human intelligence through machines. RPA is a process-driven technique, versus data-dependent AI. In other words: doing versus thinking.
High on the agenda
Cognitive RPA is a hot topic right now, as the Forrester Predictions for 2019 show. No less than 40% of all companies will develop digital employees or bots by combining RPA and AI this year. The RPA market will reach $1.7 billion in 2019 and even $2.9 billion in 2021. By the end of 2019, one in five of all interactions via the service desk will be automated by a successful combination of cognitive systems, RPA and various chat bot technologies.
Time to invest in classic or cognitive RPA?