NN is one of the pioneers in robotisation in the Netherlands. A robotisation project was launched five years ago and has since led to several hundred digital staff who assist employees by taking repetitive tasks off their hands. Rocking Robots discussed this matter with Peter Vlam, Manager Process Management & Innovation at NN.
Peter Vlam’s team is responsible for process management and innovation in NN’s Finance Service Centre. That is where the robotisation of NN started five years ago. We were getting more work as a Service Centre and wanted to tackle the challenge of doing our work better and more efficiently. Robotisation was a good solution.
RPA had come into focus at a number of conferences Mr Vlam attended. “Robotisation first got a foothold in the finance corner. We realised then that, precisely because it was all new, we needed not only software, but also an implementation partner. We then started a quotation process and talked to several suppliers.”
NN ultimately opted for the softwarerobots of UiPath. “We wanted to see whether it worked as soon as possible. Through intensive cooperation by and between business and IT, we made sure that we had our first robot running within four months that did the work of two FTE from the outset.”
“Incidentally, our main concern in deploying robots was to solve peaks in workload. It never intended to replace part of the workforce with robots. I personally believe that robotisation can contribute to making work more enjoyable and acceptable. Tedious and repetitive tasks can be done easily by a robot, so that you as an employee can focus on the more interesting aspects of your work. People who have just graduated from university also expect work to be arranged in this way.”
NN now has about a hundred robots operating by now. “And that’s just in my team, at the group level. Robots are also being used in other places within NN, so there are many more in the entire organisation. We then come to 200 robots in all.”
Mr Vlam learned a lot during this process: “We quickly saw that the robots were actually capable of taking over many repetitive tasks well and of relieving the workload in departments, but meanwhile the workload in our team was increasing exponentially. So we shifted some of the focus to the standards in and the quality of the robots’ code.
“We then hired a lead developer — a software developer within the Finance Team. This enabled us to gear all our efforts to standardisation, and thus to the quality of the robots. You can now see this in our incident graphs. For one and a half years now, the number of malfunctions in our robots has been structurally decreasing, while the number of robots supporting our employees continues to grow. This has been made possible by constant analysis of the robots’ data.”
He sees the main benefit of deploying RPA in improving the work-life balance for NN employees. “We take the peaks out of the workload with the robots in particular and make the work more exciting. Saving FTEs has never been the goal. And because we keep showing this internally, the degree of acceptance of robotisation is reasonably high. There has never been any resistance from the organisation, but explaining the goals and benefits properly does require attention.”
RPA is developing more and more as a technology, and new possibilities and applications are constantly emerging. Mr Vlam follows this closely: “There is a lot of talk about the democratisation of robotisation, and of a robot for every employee. I would like to qualify that somewhat. If everyone were to build robots that have a role in primary or supporting processes, a new risk would emerge. I do believe in partial democratisation, to use the power of the crowd and make robotisation even more accessible. However, that requires a well-adapted funnel, because you need perforce to keep the (decentralised) developed software standardised and manageable.”
“We follow the new developments offered in the RPA platforms with interest, such as test automation, for example, but we always set it against other solutions that we already have in place. And that also applies to things like process mining. That is hugely interesting, but it really only works well if your entire process is supported by a single application.”
On the AI front, NN already has an Analytics Lab that deals with things like machine learning. “That is where our data scientists are, and of course they have already chosen their own tooling in large measure and are also looking at the deployment of AI much more from a policy and ethical perspective. That said, I can see my team working more and more with the Analytics Lab over time.”
Looking back, he notes that it made sense to provide proof in the short term that it was altogether possible to automate parts of a process with a robot, and to show that it could actually help employees and departments. “In retrospect, we started thinking too late about how we could standardise robotisation and make it future proof,” he says.
“I would advise organisations starting now to do that from the outset. And it’s important to think right away about how you deal with robots as entities in your organisation. They are not applications, but they are not people either. It’s important to determine that up front, especially with respect to security and authorisation issues.”
He expects to use more and more “attended robots” in the future. “The robots we have now run for the most part in the background, but I foresee that we will be installing more and more robots on the workstations, which can be deployed more or less on demand. And above all, we will continue to optimise further the entire process for the deployment of robots itself. It is important to set a good example as a process optimisation team.”
For more information about RPA in the insurance industry, see our partner DIA TV