The robotic software revolution and mobility

Software robotics are best defined by what they do rather than the technology they use. By interpreting complex information, they respond to it and perform multi-step tasks. They are often able to ‘learn as they go.’

As well as helping companies do things more quickly, cheaply and efficiently, software robotics such as robotic process automation (RPA) applications can deliver qualitative improvements such as eliminating human errors and freeing employees to do more challenging and value-added tasks. They often improve the experience for the customer too by making processes more responsive and easier to use.

In the coming years and decades, the ability to deploy RPA effectively is likely to be a crucial competitive advantage for any firm in the mobility ecosystem that has back office functions.

One area with the greatest potential is finance and procurement, especially in order-to-cash and procure-to-pay processes. There is also RPA potential in – to name a few examples – compliance, customer service, HR, and pricing where services tend to be standardised and repetitive. Marc Engel, Partner, Mazars, describes how RPA has helped one global transportation company. “They had to check and approve thousands of invoices every month. It was dull work. But by developing a robot that could take over, a couple of full-time employees were able to move on to do more interesting and valuable work.”

Contact and call centres are another example. “Logistics companies get a lot of inquiries,” says Engel. “Where is my package? When does it arrive? Can I return it? A large express company used to have people answering these emails in person, but they noted that 80% of inquiries were pretty much the same.” They used chatbots to handle this kind of similar inquiry – freeing up human customer service agents to provide more attentive and better service for more complicated issues.

Cheaper, better, faster, stronger

Martin Váross, Partner, Mazars, says there are broadly two types of approaches to identifying tasks that can be automated in this way. “The first involves examining standard processes, which our team analyses. We look at repetitive or low value-added activities that meet certain criteria based on our experience – those that require greater work effort or have higher error rates.”

The second type, he says, is assessing how employees or users go about their operations day to day, with a view to streamlining them - a process known as task mining. “We recently did this for an insurance company,” Váross explains. “We analysed the user data in two departments that handled information from health providers. By understanding how employees were doing reconciliation checks, we found opportunities to reduce the number of screens they needed to use, automate tasks, and find other efficiencies.”

Engel adds that companies are often surprised about how quickly and cheaply processes like this can be automated. “They expect it to cost hundreds of thousands of dollars,” he says, “when in fact it's closer to tens of thousands. People expect it to be a long, expensive IT project. It's often simpler than that, and they can start to deliver savings quickly.”

How to prepare for automation

Engel and Váross say the experience of working with many different companies provides a clear picture of what firms need to do to implement software robotics successfully and reap its benefits, citing three key lessons.

The first is the importance of selecting the right processes for automation. “Companies whose RPA programmes fail normally do so because they pick the wrong processes,” Váross says. “We advise companies to automate processes that are repetitive and relatively stable. Expectations have to be managed very carefully so that everybody understands which parts of a process can be automated and which can't, and how to combine RPA automation with full-scale automation.”

Second, companies should always start by automating a lower-value, simpler process, to build some understanding of software robotics into the firm. Only then should they try to scale up their use of software robotics.

Third, Váross says companies that integrate RPA processes most effectively tend to be ones that have set up dedicated RPA governance functions to provide clarity on responsibility for identifying issues, troubleshooting, and making sure processes continue to improve once they have been put in place.

The sky is the limit

Engel says that he has noticed the Covid-19 pandemic push companies to speed up their digital transformation process. “Clients are digitising faster and more broadly than pre-pandemic,” he says, citing the example of Dutch municipalities that adopted RPA technology to quickly and thoroughly handle peak volumes of pandemic-related benefit assistance claims from companies. “I expect the pandemic will increase digitisation across the board,” he adds.

In the long term, RPA within the mobility ecosystem will be as varied as imagination allows. By collecting more detailed data on carbon emissions, for example, transport apps might be able to automate emission calculations for different routes, helping travellers make greener choices. Within the next decade, driverless vehicles could become more common, powered by automated software to process the data from the road, other vehicles, the weather, and passengers and then synchronise it all to keep the vehicle operating as smoothly as possible.

In the years to come, software robotics are likely to go from streamlining the back office functions of original equipment manufacturers, airlines, logistics companies and public transportation authorities, to enabling totally new forms of mobility. Once that happens, the sky is the limit for how these robots will revolutionise how we travel.