Whilst you may be familiar with the term, “Robotic Process Automation”, or “RPA”, which we delve into in another post here, you may have heard the term, “Intelligent Automation”. You may be wondering whether RPA and Intelligent Automation are two labels for the same thing, or, if there is a difference?
On one hand, RPA mimics rules-based tasks and processes. However, chronologically, it often forms the second of five key components that comprise Intelligent Automation, which we have broken down here:
Component 1- Business Process Management (BPM)
BPM involves the mining, defining, mapping and monitoring of business processes. Essentially, when end-to-end BPM software, like Signavio, is implemented by its partner, Venturiq, it looks for, defines and monitors processes. This can be used to identify processes that are ripe for automation and to check on the efficiency of automation. As such, well thought out BPM acts as a base layer to all automation. Hence it is submitted that BPM is the first component of Intelligent Automation.
Component 2- Robotic process Automation (RPA)
The basic idea of RPA is to copy and repeat business processes that follow pre-defined business rules. When a coherent implementation model and architecture, such a Blue Prism’s Robotic Operating Model, executed by a certified partner like Venturiq is in place, it sets the RPA program up to morph into Intelligent Automation. More on RPA, the second component of Intelligent Automation, here.
Component 3- Artificial Intelligence (AI) & Machine Learning (ML)
AI is when machines are used to simulate human intelligence. More specifically, AI is a system that is capable of learning and reasoning, bringing the ability to make judgement calls, such as by reading text and interpreting images. AI is made up of many smaller concepts and capabilities. For example, Machine Learning (ML), where patterns are studied to inform decisions, Natural Language Processing (NLP), Computer Vision, such as ABBY’s leading Optical Character Recognition software, implemented by Venturiq, Big Data and more. Some examples are-
- The ability to decipher patterns from data
- The ability to make decisions intelligently
- Predictive and prescriptive analytics
- General Improvements to user experience
When the ability of AI to make judgement calls, as well as the possibility for computers to learn and improve, through Machine Learning, inform and augment the process steps that Digital Workers then follow, it is the third component that turns pure RPA into Intelligent Automation.
Component 4- Integration
Many companies these days will have multiple systems and interfaces for different departments and it is all too common that these systems struggle to work in a coherent manner. Every system can cause a different problem and historically many hours in the day are wasted duplicating work or fixing compatibility and integration issues. Generally, companies will acquire Application Programming Interface (API) software which functions as a bridge between these clashing systems and will help ease the burden placed on employees. Hence, the ability of a Digital Worker to execute a single process whilst switching between different systems, platforms and programs (whether through robotic log-in or API connection) is the fourth component of Intelligent Automation.
Component 5- Interaction
With pure RPA, a Digital Worker is left to complete a defined and strict process-driven task from end-to-end. However, there are two types of “interactions,” which help turn RPA into Intelligent Automation.
The first involves interaction between different Digital Workers, when disparate RPA processes, as well as distinct Digital Workers trained to complete separate processes are linked to each other, with well-designed orchestration and scheduling. In essence, the tasks executed by one Digital Worker help other Digital Workers to complete their seemingly separate tasks.
The second interaction that turns RPA into Intelligent Automation, is when Digital Workers interact with human colleagues. This could be for exception handling, where only human judgement will do, or it could be temporary, during the “training phase” of Machine Learning. Effectively, this interaction, which is otherwise known as “human-in-the-loop”, enables Digital Workers and Human Workers to interact like colleagues.
Open for Debate?
So those are Venturiq’s thoughts on what turns RPA into “Intelligent Automation”. What are your views? Does the above analysis succinctly cover all components of Intelligent Automation, as characteristically different to RPA?