They say, health is the most important thing, because if you haven’t got that, you haven’t got anything. And, thanks to technologies, like Magnetic Resonance Imaging (MRI), detecting diseases can be done quickly, accurately, and without any side effects. But, that still leaves doctors with the question of finding the best way to treat what they find?
The same situation applies to businesses. After using tools like SAP Process Mining by Celonis to perform the equivalent of an MRI on the enterprise, how do you go about dealing with the deviations and hidden inefficiencies in the processes that lie at the heart of any operation?
The stock answer is “automate them!”
Different strokes for different tasks
Automation improves the user experience for both employees and customers. RPA boosts productivity and can typically free up between 15% and 30% of an employee’s time to focus on tasks that deliver greater value to the organization and improve the customer experience.
This simple and nonintrusive adoption of advanced technology in an employee’s day-to-day tasks enhances employee well-being and satisfaction. It can also make an important contribution to a company’s change management as it seeks to accelerate its digital transformation.
Different types of RPA robots are available to help with a variety of tasks. The mostcommon types are Attended RPA and Unattended RPA.
Unattended RPA includes processes that can be automated end to end by using robots installed in server clusters. Attended RPA interacts with a desktop application, with the robot acting as a software assistant for a human user.
Fast, but not always flexible
The deployment of Attended RPA can be very fast, because it has no impact on existing information systems and requires no modifications to applications, which continue to function unchanged. As a result, attended RPA projects are short and ROI is rapid. It takes only a few weeks to set up an RPA robot that can save a significant percentage of time for tens – or even hundreds – of employees.
But, that’s not necessarily the whole story. A recent article in Forbes warned that businesses should think twice about RPA, and outlines some potential problems. For example, RPA is especially useful when interacting with older, legacy applications. But, it isn’t always flexible. If anything changes – such as the interface, the data, or the legacy app – then RPA can run into difficulties. The robots don’t adjust to new conditions like a human being might.
Add AI to create intelligent RPA (iRPA)
That isn’t to say that enterprises should avoid RPA. Far from it. RPA robots can make business processes more efficient and employees more productive by taking advantage of structured data stored in various databases and application silos. But what about the increasing number of business scenarios involving unstructured data such as images, text, and speech – often coming from mobile devices?
For these types of specialized cognitive tasks, the answer is artificial intelligence (AI). AI has advanced rapidly in recent years. Thanks to the huge amounts of data now available to train the models, machine learning and deep learning algorithms have reached very high levels of confidence, beating human cognition in many cases.
Providing RPA robots with AI capabilities could deliver unprecedented levels of business speed and efficiency – and it doesn’t stop there. Within a few years, AI could help to bring self-learning capabilities to RPA robots. By understanding what a human user is doing, bots could replicate some tasks and even adapt themselves to minor changes in their working environment, such as handling exceptions or updates in the applications they interact with.
Step by step
Automation cannot simply be inserted into business processes and operations without considering the wider impact. There are some key factors that are worth considering the right steps.
- Understand the bigger picture: Automating a specific process can have implications for other related processes.
- Refine first, automate second: RPA automates processes, but it doesn’t necessarily improve them. Before considering automation, evaluate how the process might be refined – particularly by adding intelligent capabilities such as advanced analytics or machine learning.
- Prepare for the future: It’s vital to document and catalog each bot’s functionality and how its code works. Business process innovation is constantly accelerating, and the bot may need to be enhanced with machine learning or AI capabilities in the future.
- Have a plan: Whether planning to implement across the organization or start with small pilots, it is essential to have a strategic plan. For larger enterprises, it may make sense to create an RPA center of excellence to act as a repository of policies and strategies for RPA deployment, and catalog bot capabilities and dependencies and reskill the workforce.
Want to know more? Register for our upcoming webinar Diagnose & Improve Enterprise Health with SAP Process Optimization, live on 12 Feb at 10:30a.m. Hong Kong Time.