RPA and Artificial Intelligence, or process performance management
RPA has become the trendy tool for optimising processes in organisations in a short period of time, but there is no miracle tool (yet!), and process management is a broad topic that brings together several somewhat different technologies.
When we talk about processes, the first word that comes to mind for business or IT departments is often ‘modelling’.
Here we enter the world of BPM (Business Process Management), which appeared in the last century with its historical tools, its international standard born of several years of reflection, the BPMN (Business Process Model & Notation), and its recent developments. The arrival in force of RPA should not make us forget that BPM is still relevant. The automation of processes does not prevent modelling efforts. Recent BPM tools also provide collaborative functionalities, ‘real-time’ process monitoring, management of alerts, documentation, profiles or resources concerned, and even optimisation solutions.
Then it’s just a short step to the world of BPA (Business Process Automation). As the name suggests, processes are automated here. Nevertheless, at this stage, the tool will allow predefined jobs to be launched (or programmed for the needs of the cause) without being able to easily take charge of all the actions present in a complex process, nor integrating a user interface capable of modelling the actions to be carried out by the employees; this will justify the introduction of RPA. However, with this major evolution, a fundamental architectural concept is incorporated, which allows BPA to be integrated into the IS and to connect, by means of the necessary APIs (application interfaces), with any other solution in the system. This aspect will of course be found in RPA solutions.
Here we are, at last, at the heart of the matter: What does RPA add? Robotisation reproduces the process tasks as they are performed daily by an employee. Of course, BPA can do this too, but the associated job must be programmed. RPA has an interface that makes it much simpler to do this, either by capturing the user’s gesture or by chaining together pre-programmed ‘robots’ provided in a library of the solution. Both technologies have their advantages and disadvantages, but both provide the ability to robotise any process. It is in this user interface that the originality and added value of RPA lies. It is relatively quick to learn and remains accessible to less technical profiles (with a certain rigour in the reproduction of processes, and a limited complexity of task chaining rules).This brings us to the point where human actions are potentially replaced by RPA. But human replacement means ‘Artificial Intelligence’ (AI). First of all, it should be noted that not all human actions will be replaced.
The interpretation of data in a complex context, the transformation of a process caused by an external event, the decision-making linked to the commitments and responsibilities of a manager … are all examples that require human reflection, and that will not be replaced in the short term by the robot. Nevertheless, advances in AI are now enabling a number of advances, ranging from simple text recognition (OCR) to deep learning. The purpose here is not to define the terms and scope of AI, but to understand its interest, in the context of process performance management, and in association with RPA.
OCR or NLP (Natural Language Processing) are analysis tools that enable performance improvement. Chatbots provide a first ‘intelligent’ answer because they can communicate with humans. But it is in Machine Learning (ML) and Deep Learning (DL) – which are also associated with the recent Chatbots – that AI brings its full power to process management: decision support.
The aim is not to replace humans, but to advise them on how to behave in a situation where they have to make a decision. The latter solutions rely on Machine Learning to improve over time and on Deep Learning to understand situations by correlating information that they already master.
This is the ultimate stage of organisational process performance, at least as long as we decide that people still have a role to play in the functioning of our society !
How do you get started with process optimisation?
The objective is not to multiply the tools, quite the contrary, but to equip oneself with the appropriate solution for one’s immediate needs and for developments over two to five years. The impact on the company’s organisation must be taken into account, as well as the capacity of the systems to provide the right information at the right time. This brings us to the no less important subject of data governance, which is closely linked to process governance and quality monitoring. We are thus implementing good practices, and a governance that does not imply starting the whole Quality approach from scratch. The implementation of RPA is often an opportunity to optimise internal operations, little by little, process by process.
Manager in charge of the RPA offer.