Understanding robotic process automation, artificial intelligence and intelligent automation
Robotic process automation (RPA), artificial intelligence (AI) and intelligent automation (IA) are recurring terms in events, specialized and generalist media since the last three years. There is undoubtedly an increasing hype around these subjects, but some companies and people are using it as buzzwords, without fully understanding it, which leads into a lot of confusion. Here are short definitions to grasp their meaning:
Robotic Process Automation (RPA)
It is important to keep in mind that RPA has nothing to do with physical robots, and, in its basic form, not with artificial intelligence either. RPA is a software application that replicates the actions of a human being interacting with the user interface of a computer system – and subsequently completes tasks automatically and repeatedly. In fact, RPA is basically a set of rules-based algorithms programmed to perform one or several specific tasks. Each of these algorithms are commonly referred to as ‘bot’ or ‘robot’. The attraction is that it is -perceived- quick, non-invasive and simple to deploy, while promising significant cost savings achieved within a short timeframe. However, experience has shown that, only because it does not require any programming skills or IT integration between different systems, it still requires planning and preparation, hence is not as easy as widely claimed. Moreover, companies have to be clear that RPA alone is not ‘intelligent’, which means it does not learn from experience and cannot automatically improve itself. Each change or upgrade needs a human intervention.
Artificial Intelligence (AI)
AI is by far the most complicated of the three technologies because it covers several different techniques and can impact all sectors in many different ways. Because of that, there is no consensus on a definition of artificial intelligence. Yet, we set a definition trying to be as inclusive as possible: AI is the combined use of algorithms, knowledge bases (big data sets), and neural networks/deep learning techniques to mimic and complement human abilities in a variety of domains, including:
- Perception and understanding
- Reasoning and problem solving
- Learning and training
- Interaction with surroundings and people
AI algorithms, unlike RPA, are supposed to be ‘self-learning’, which means they improve their results through training. Thus, fed with more data, they can become more efficient over time. However, there is still a lot of research required in all fields of AI, while RPA a is today a more complete technology. For example, “strong AI”, an AI conscious and able to realize multiple tasks does not exist today, and might not exist for decades.
Intelligent Automation (IA)
This term refers to the use of RPA combined with some AI techniques. An example is a chatbot based on AI algorithms: some of these chatbots can trigger an RPA bot at some point in the ‘discussion’ with the user. In this case, there is a distinction between the chatbot powered by AI algorithms and the RPA bot that follows a process ‘called’ by the chatbot. So, intelligent automation is a kind of RPA which is built with an algorithm able to automatically solve specific issues in the process it is supposed to follow. The main RPA software vendors, like UiPath or Automation Anywhere, and IT services companies, like Accenture or Wipro, are working on Intelligent Automation solutions, which have many advantages such as better efficiency in the automation of tasks and the ability to detect and solve problems faster than humans.
To conclude, RPA and AI are two different technologies because their respective algorithms are built with very different techniques and intentions. While RPA focuses on the automation of processes defined in a computer system, AI has many more possible applications. However, the combination of RPA and AI is possible, resulting in ‘intelligent automation’, which still focuses on processes, but can do it in a more efficient way thanks to AI.
Subscribers to PAC's SITSI research wanting to know more about RPA, AI and IA, the trends and main actors, you can read our report Robotic Process Automation, Artificial Intelligence and Intelligent Automation – Status Quo and Key Trends or contact us.
For a full understanding of artificial intelligence, you can read this report: Artificial Intelligence: Demystification and Market Opportunities