If I asked you to visualize robotic process automation (RPA), you’d perhaps visualize thousands of bots working over-time to perform menial tasks. Something like this, for instance.
[IMG: Thousands of bots working inside computers]
I see why. Since the beginning, RPA has been sold and bought as the cheapest way to scale the workforce of an organization to perform high-volume tasks.
Do we have a million invoices to parse through?
Do we need to copy billions of files to back up?
Do we need updates sent to customers every single day?
Do we need to scan gargantuan Twitter feeds for keywords?
This brought measurable improvement in the productivity and performance of the workforce in general. In fact, even today, RPA performs such tasks world over. And with good reason.
But then again, employing thousands of bots to individually perform menial tasks is a crying shame given the advancements in cognitive automation today.
Round-the-clock bots for when you’ve called it a day.
The most recognizable RPA is the personal assistant, say, Siri or Cortana. These are what we call ‘attended RPAs’. They respond to a specific trigger by a user and make their life easier. Their usefulness doesn’t necessarily end at making a calendar entry or playing a song. They could even perform more complex tasks such as scraping for information or copying it from one place to another. However, they rely entirely on user prompts. So, they work the hours the user works.
Unattended RPA, on the other hand, can execute workflows without human intervention. By being independent of the user, these bots can work 24×7 — do more, achieve more. Thus, you’d need fewer bots.
Multi-tasking bots that are always productive.
Modern bots are master multitaskers. Unlike its predecessors, who were programmed to do a single task over and over, today’s bot can switch between different tasks with minimal reprogramming. As and when a specific task is complete, these bots will have the intelligence to understand what are the next tasks in the queue and proceed to complete. Of course, they can always return to the first task later if necessary.
Having multi-tasking automatons means that you don’t need a ‘bench strength’ of bots, for when additional work comes your way.
Pre-trained bots when you need an experienced hand.
RPA solution providers, with deep skills in niche areas and intelligent automation capabilities, are coming out with bots that are pre-trained in specific processes. Instead of building bots that require frequent updates and take 6 months to configure, automation companies are building specialized bots that can be deployed in less than 3 weeks. For instance, insurance companies can go to a vendor, buy pre-trained bots for risk arbitrage and have the bots on the task before they’ve even returned after signing the deal.
These bots are what a hiring manager would call lateral entry — a mid-management, independent, self-starting bot! Not only does it save on implementation time and resources, but also works at top standards with little data or human help.
Truly intelligent bot when you need a loyal workforce.
The real impact on ROI, productivity and competitive advantage comes with cognitive RPA or more popularly known as intelligent automation. Cognitive RPA is one that combines speech recognition, natural language processing (NLP) and machine learning (ML) to mimic human behaviour — making judgments and decisions based on intuition.
With every new data point that the RPA encounters, it makes new connections, learns and grows, unsupervised. A cognitive automation solution gains ‘experience’ the way employees do. See what the cognitive RPA at Mondi is doing, for example. Mondi is a global packaging and paper company. They are using RPA to “collect data on the shop floor system, bringing it back to a machine that then learns from it and improves the product at the end of the day.”
As these bots spend more and more time with the organization and its customers, they learn and adapt the way employees would. One might even say they do it better. They combine the best of both worlds — the efficiency of automation and the empathy of human interaction. Therefore, organizations no longer are restricted to a one-trick-pony, nor do they have to shed millions of dollars to buy thousands of one-trick-ponies.
Are you ready to fire your starter pistol?