Robotic Process Automation (RPA) is beginning to gain momentum in the enterprise. RPA automates repetitive, structured tasks and can manipulate data, communicate with other systems and even process transactions. And the market for RPA is expanding as enterprises look to trim costs and boost efficiencies.

The Institute for Robotic Process Automation recently predicted that RPA will change organizations' approach to business processes and IT support, helping to streamline workflows and strategies surrounding remote infrastructure and back-office work. 

In a conversation with CIO Dive, Dave Moss, CTO of Blue Prism, discussed how enterprises might handle the rise of the bots and how the IT organization can make a smooth transition to RPA technology. This conversation has been condensed. 

Why would IT be cautious or hesitant about RPA solutions?

Moss: At first glance, RPA could appear to the IT folks to be nothing more than scripting and screen scraping—which are technologies that are not enterprise-strength and create more problems than they fix.

For RPA to work at an enterprise level, it has to behave as an enterprise piece of software, which means no record buttons or downloadable versions to "play with." You can't play with enterprise software; it's serious stuff. IT needs to apply the same level of rigor and investigation to RPA software as they would to any enterprise software.

How can IT tell which solutions are enterprise-grade?

Moss: IT already knows how to evaluate enterprise software in terms of proven references at scale, operational resilience, transaction integrity and recovery. When it comes to evaluating different RPA options, though, IT's job becomes harder because there are many solutions out there putting themselves in the RPA category, but most of them can’t operate according to enterprise requirements.

RPA needs to be centrally hosted in the data center, not on individual computers and desktops, in order to give IT the security and control it needs. For example, our customers are processing millions of transactions a month through our RPA platform with virtually no human interaction other than to deal with the occasional process exception. It needs to function this way at this scale of transactional work because the application is too big to fail: it’s a mission-critical application.

When IT thinks of enterprise deployment of RPA, it raises the bar from simplistic, small-scale desktop-deployed business support to high-availability data center mission-critical transaction processing. They are very different: one is strategic while the other is tactical.

Why does IT need to be involved in the deployment process from the beginning?

Moss: To support enterprise levels of availability and security, IT has to set high standards and stick to them if the RPA platform is going to deliver continual business value for mission-critical processes and transactions.

Some organizations drive this kind of implementation from the business operations team, which could leave IT behind in their rush to get the deployment up and running and leads to "shadow IT" being installed. Involving IT from the beginning ensures that it has the chance to make sure the solution and deployment is fully compliant with enterprise standards.

What are some steps or tips for implementing and managing a virtual workforce alongside human employees?

Moss: For a smooth transition, it's crucial to have employees embrace the change, rather than fight it. Communicate early and often. Involve employees early in the process and give them ownership of the virtual workforce; after all, they're the ones who will be treating the robots as coworkers.

Focus on transitioning the mundane and repetitive tasks that your employees don't want to do; people aren't robots. We've seen companies where the human employees embraced their robotic coworkers, giving them names and faces, and asking if they could be trained to do even more work for them. If it's framed properly, people will embrace the change.

Could RPA eventually prove the end of the need for certain kinds of workers?

Moss: It could reduce the number of mundane, "robotic" jobs, as technology has always done; for example, you don't see many filing clerks or typing pools any more, nor are they jobs that most people want to do.

RPA has much greater potential to transform jobs, as it would take the repetitive, programmable tasks off people's desks and free them up to do more of the creative, problem-solving, value-add parts of their jobs that can get lost in the shuffle now. It also has the potential to create different jobs, such as robot technicians.