The New Wave in IT - RPA & Cognitive Bots

We have been reading and listening a lot on automation, robotic process automation and cognitive “bots” (consider the "bots" as software equivalent of robots) these days. Though there is no clear indication as to which assets falls under which category, I would like to put forward my ideas for a general layman understanding of these keywords.

Automation is the application and implementation of technology so that any job, task, process etc. can function automatically. Automation makes life easier, convenient and increases efficiency in day-to-day operations. Integration of systems, automating workflows so that certain pre-defined actions gets triggered at each level, improving the speed with which business processes takes place. Automation is technology dependent and requires complex integration. For e.g. the automated meters helps consumers when they fill-in gas. Or the time when you had to stand in queue for a long time to withdraw cash, that process has been automated via ATM machines.

Automation, as you can comprehend, is a very wide term. We can classify them as robotic process automation(RPA) or cognitive automation

Now let’s understand robotic process automation.

Think of any business or IT process which was earlier done by humans, for e.g., consider filling up 40 fields in an order placing form. The steps in the order placing process involves copying data from one of the standard templates and pasting it in another standard template. As you would have understood by now, this process is pretty straight-forward and does not require any knowledge worker. Hence, if in some way, we can write a piece of code which can read the field from the source and paste it in destination template, we have achieved automation!

So, in general, robotic automation. And hence, no human intervention is required for that job.

Having set the ground for RPA, let’s now understand cognitive automation.

Cognitive automation allows the “bots” to make decisions based upon inputs, just like humans do. It requires training before deployment as to what it has to do. For e.g. consider the same copy-paste example. From the standard input template, the bot copies the data, but now the output (pasting template) forms’ language i.e. the language of the order placing depends upon certain field of the input template, and hence automation, in this case can be achieved by cognitive bot. The Cognitive bot will translate the whole output form in the required language and then finally place order.

Now, let me take another example to distinguish robotic and cognitive automation. Consider a job which involves putting cashew nuts at the center of hand-made biscuits.

Case in point: Robotic Process Automation

Here the shape of the hand-made biscuit is fixed and hence the “bot” knows where is the center of the biscuits and hence can put the cashew at its center.

Case in point: Cognitive automation

Interestingly, now there is a flexibility on the shape of the biscuit, implying the shape can be circle, rectangle, heart etc. Nevertheless, the “bot” figures out the shape of the biscuit and makes a decision as to where the center of that shape is and puts the cashew there.

Given the current market scenario, it is imperative that you understand what is the underlying definition of RPA vs Cognitive bot and what capabilities you can expect out of them.