And how does it help with everyday tasks?
‘Automation’ describes an action that’s conducted independently of human intervention.
Early forms were mechanical or, later, electrical. We’re already in the age of computer-enabled automation–and exciting new forms, enabled by cognitive computing, are on the way. The term Robotic Process Automation (RPA) may sound as though it’s about replacing people with androids–robots that resemble humans. In reality, the ‘bots’ in RPA are more like the Macro facility in Excel.
These bots are software that can span different systems.
This means RPA can automate many routine ‘swivel-chair processes’, where people take data from different systems to be combined, analysed or simply actioned. So RPA is not about androids taking human jobs. It’s about software bots performing many of the routine processes currently performed by highly skilled knowledge workers–such as accountants.
In short, it frees professionals from laborious tasks and allows them to focus on adding value.
RPA is straightforward to implement. Because it sits on top of current systems, it can be deployed without re-programming or making any changes to current systems. In fact, the proof-of-concept stage of an RPA project can be completed within just a couple of weeks, and it can be rolled out a month or two later. So RPA is a relatively inexpensive way of providing work-arounds and making legacy systems more efficient. But be carefulit can also mask the need for an upgrade, weakening the business case for a major systems replacement initiative.
RPA is not about androids taking human jobs.
RPA is a simple form of automation. That makes it a good starting point for those at the beginning of the automation journey–bots follow instructions and perform tasks as instructed, faithfully replicating a human’s actions.
But they’re not perfect. They have to be kept up to date, and a person with a good understanding of the process will need to take over any transactions they cannot handle.
RPA is a relatively inexpensive way of providing work-arounds and making legacy systems more efficient.
Less labour time
Fewer errors
Easier fixes
Works with current systems (no need to ‘rip out and replace’)
Relatively inexpensive
Speed (proof of concept in two weeks, working inside two months)
Good governance is vital (you need to know where the bots are buried).
Bots need to be maintained or amended: they work best alongside people who are familiar with the processes.
RPA can delay recognition of the need to replace legacy systems.
