A guide into the world of intelligent process automation

24.4.2018

The field of robotics is full of excitement and buzz-words, but it is often difficult to determine the concrete under the hype. The term Intelligent Process Automation, more commonly known as IPA, has recently joined the conversation. What is it all about? And why is it worth paying attention to?

What does IPA stand for?

IPA refers to a process automation solution, where the technology being used is smart – at least to some degree. There isn’t just one kind of IPA, rather the solution is tailored to fit the process requirements. The technologies utilised in the solution might include for example software robotics, chat bots, image recognition or machine learning. Put short, IPA is an umbrella term for a variety of different technologies that can be utilised (together) to automate processes.

IPA as a continuum of virtual work

It is reasonable to consider a company’s shift to using digital workforce as a continuation, where the journey starts at simple, repetitive and easily defined tasks. These routines of knowledge work are automated cost-efficiently with RPA (Robotic Process Automation), where every action of the robot is predefined in a manner following “if this then that”-logic. RPA best suits a situation, where different steps are repeated predictably and volumes are high – this way savings in expenses and the quality of operation rise enormously! RPA that is based on predetermined rules however doesn’t fit a situation, where steps can’t be well defined because there exists a large variety of possible actions following a situation.

In the next step of the continuum, focus shifts towards smart technologies, that take up tasks that require interpretation. To understand this step, it is useful to consider an RPA exception situation, where a software robot operating based on predetermined rules sends back a group of exceptions for the human worker to deal with. In practice, a situation like this might occur for example when a specific piece of information is logged in the used IT system’s open field in different formats. When the software robot processes the information, it only recognises the predetermined formats and reads the differing formats as errors. In a situation like this, a virtual worker utilising machine learning might provide a solution, by learning to identify that the different formats have the same meaning. If the smart virtual worker is delivered as a cloud-service, the solution also enables a flexible level of service, where more capacity and tools may be put to use when necessary.

Artificial intelligence in turn refers to completely autonomic systems, that can interact with their surroundings at any situation and reach their goals independently. These kinds of technologies are represented by for example IBM’s Watson and Google’s Alice. As computers’ abilities grow it becomes easier to recognise what technologies can’t be considered in terms of artificial intelligence, though defining true artificial intelligence remains difficult. For example, photo recognition – today considered to be routine technology – was previously considered a form of artificial intelligence.

Why is understanding the big picture so important?

The conversation around robotisation and future of work is often regrettably vague and leaves it unclear, what are the practical applications of the new technologies. Pumped up by the hype, it is easy to start doing things that sound great but do not meet needs in practice – at least not cost-efficiently or in a way that delivers on expectations. On the other hand, because the change in all fields only keeps accelerating, and because the competitive advantages reached by utilising virtual workers have been unprecedentedly large, this boat should not be missed!

The utilisation of virtual workforce should however always begin by considering the need, in a way that the choice of technology happens on the terms of the target. The best results are reached by climbing up from root to the top: transforming business processes into a digital form that can be utilised by robots and moving from rule-based automation towards smart solutions where the needs are clearly recognisable. The implementation of smart automation typically starts by recognising and prioritising the objectives. The concept of ‘IPA-continuum’ represents this way of thinking.