Robotic Process Automation and Intelligent Automation - a primer

Introduction

Robotic process automation (RPA) generally focuses on automating repetitive, frequently rule-based activities, whereas intelligent automation (IA) uses artificial intelligence (AI) technologies, including machine learning, natural language processing, structured data interaction, and intelligent document processing. While we often times use these two terms interchangeably, there are key distinctions between them. This article will cover this with some simple examples.

Automated and intelligent technologies are primary contributors to the next phases of digital transformation. These technologies are not only transforming the way that businesses operate, they are also redefining human work as well. The speed with which tools such as robotic process automation (RPA) and Intelligent Automation (IA) are being integrated into all phases of business operations across all industries is staggering, and the momentum is here to stay.

What is RPA ?

RPA is a software technology that transforms the human aspect of work that is routine and repetitive. While we may not realize it, traditional work that we do all the time manually and repeatedly involves a lot of tedious tasks throughout the day, such as:

• Opening email and attachments
• Logging onto multiple websites to extract information.
• Copying and pasting data amongst multiple spreadsheets
• Collecting statistical data
• Reviewing invoices and putting them into billing system
• Moving data from one system to another
• Copying data from here to there

RPA bots, for example, can reside on one’s desktop and can perform its tasks using the same employee ID as the human worker it’s assisting. While it can operate as a virtual assistant, it can also operate on its own (attended vs unattended model of RPA deployment, or hybrid, based on use cases and needs). It can open a web browser, type in a URL, log on with supplied credentials and extract the required information from it to input into a document or spreadsheet.

What is IA ?

While RPA simulates human work actions, it doesn’t simulate human intelligence. For example, it can download data and transfer it to its desired location, but can’t interpret the data and draw any conclusion from it. It only operates under the rules that define its actions.

On the other hand, IA operates more like a complete human because it can interpret data to make inferences and conclusions from it. For instance, Intelligent automation is being used in logistics today to streamline supply chains by identifying potential bottlenecks before they even happen. This predictive intelligence helps it get the right assets and resources in place ahead of time.

IA utilizes multiple technologies to help it work intelligently, including RPA. Of course, as its name implies, IA utilizes artificial intelligence (AI) to help it simulate human intelligence. This helps it analyze data far faster than any human can. Thanks to advanced algorithms, it uses machine learning (ML) to help identify patterns in large volumes of data. Other incorporated technologies include computer vision tools such as optical character recognition (OCR) that convert scanned documents or photos into text. Natural language processing (NLP) is used to communicate with humans through a conversational interface, and process mining is used to diagnose business processes to improve upon them.

One example of how IA goes beyond the rules-based limitations of RPA is the advancements in chat boxes. Most online users are familiar with these tools now as companies use them to interface with customers. If you’ve ever become frustrated by a chatbot, it’s probably because you needed some information or action it wasn’t programmed for. While RPA chatbots are great at directing you to a specific web page, getting you to the right human support specialist or answering frequently asked questions with template answers, they can aggravate someone needing assistance at a deeper level. AI-driven chatbots can better interpret what a customer wants and can provide service at a more innate level.

What is ONE key difference between RPA and IA ?

One thing that makes us human is the ability to improve our level of performance over time. While a RPA can perform tasks faster and more efficiently than a human, it is incapable of learning how to improve upon itself. AI-driven bots, called IA, can learn and adapt to data and events in real-time by making tiny adjustments, allowing them to adapt to changing environments. This ability to improve and adapt makes it far more human-like than RPA.

Tools and Technologies:

There are a plethora of both commercial and open source tools and technologies that are being widely used and deployed at scale. Common ones include UiPath, Blue Prism, Automation Anywhere, Robot Framework, TagUI, Taskt, etc. One can even employ custom solutions on AWS using their Lambda and Step function. Choices are unlimited. But one caution is - perform a top down process analysis, and then determine the technology, not the other way. Remember - a fool with a tool is still a fool.

Key Takeaways:

In the same way that the web has gone through multiple generations, IA can be considered a more advanced generation of RPA. In the same way, Web 2.0 and Web 3.0 signified evolutionary steps that created advancements in how users utilized the web, AI gives organizations the ability to achieve bigger and better things than RPA was ever designed to. For those instances that require a deeper level of understanding or the ability to color outside of the lines, IA is the next evolutionary step.