What is RPA for?
RPA is used for general knowledge process work in which high-volume and highly transactional processes are performed. RPA is already commonly used in numerous operations, e.g. in payroll, quote-to-cash, onboarding, creating mass emails and faster order processing. By eliminating manual intervention, it can bring significant improvements in quality, accuracy and time savings.
Software robots enable more time for serving customers
Our first RPA robot called Robin was launched in November 2017. Since then, two more robots, Ruby and Ralph, have joined Metso. The robots currently handle around 90 different business processes in several functions.
All our robots have a manager and an identity. Their manager defines their workload, monitors their work and the results. Unlike other employees, robots don’t have sick-leaves, weekends or vacations – they can work tirelessly, 24/7. All the work that is given to the robots is scheduled so that they know what to do and in which order.
The success rate of the robots is around 97%, which means that they don’t make errors very often. Because they always review the process that they are executing, it leads to a leaner process. Also, our businesses are constantly planning new ways to utilize the robots.
The most important thing about RPA is that it enables employees to spend more time on more important tasks, such as focusing on value-adding activities and serving our customers better.
We enable smarter ways of working by using AI
In April 2019, Sara joined our other software robots. Sara is a chatbot-style virtual assistant working with a predefined task and providing answers based on a predefined list. Most chatbots utilize use cases that define a set of possible interactions between the system and the user, but Sara utilizes cognitive search.
Sara is constantly learning new information because it is utilizing AI in the cognitive searches. It has a discovery module integrated in it, so it can search for information from all the structured and unstructured data from databases, such as Intranet. Therefore, it can provide support to any query employees might have. It does this by using natural language processing (NLP) algorithms that analyze written phrases for their contextual meanings and retrieve relevant answers from one or more knowledge sources. This kind of search differs from a normal search feature in that there is an option for employees to provide feedback after each interaction. If the answer was helpful for the user, it gets higher in ranking. The answers that the user didn’t find helpful, get a lower ranking and will eventually disappear if the same feedback is given many times in a row.
The future of RPA and AI in Metso
We are currently working on improving the quality of query results, reporting functionality and some additional features that will assure better usability and scalability across support functions in the future. Eventually, Sara could potentially develop into a more mature knowledge discovery platform for complex queries that might have several possible relevant answers, a platform that employees can use to find more information on past projects, service incidents, root cause analyses, lessons learned etc.
Leveraging an artificial intelligence, chatbot-style virtual assistant is the first step in Metso’s digital strategy of AI-enabled smart ways of working. In the future, it will most likely continue to free up even more time for analyzing and development work.