Six step fix for preparing processes for automation
Before automating a process, PEX practitioners should ensure the process itself is sound. In this article, Einar Michaelsen gives his six step checklist to get processes ready for an RPA takeover.
Add bookmark
Businesses recognize that RPA alone is providing tremendous value and ROI. With the frontiers of artificial intelligence being explored, and vendors offering AI solutions to companies, it has become clear that combining RPA with AI can push this value to the next level.
Standalone RPA may be thought of as a replacement for human hands: it provides fantastic data extraction and entry capabilities and unlike humans, can carry out tasks accurately, at speed, without needing to take breaks. But it remains a pair of hands.
Adding AI to the equation means adding eyes, ears and brains to your tool-set.
AI add-ons will expand the automation potential, giving bots the ability to perform tasks that would be far beyond RPA alone, but it should not be thought of as a magic fix for your processes. Before starting to automate the process, I would recommend having a holistic end-to-end approach to fixing your processes whether you are using RPA or RPA and AI together.
This is my six step fix for preparing processes for automation.
- Refocus
Remove unnecessary activities or steps by addressing the underlying root causes. Remove waste in the process, as well as hurdles for automation, such as non-digital parts of the process. The lean six sigma toolkit is your friend at this stage as you explore what is really happening within your processes and start to understand what needs to change.
READ MORE: stay up to date with the latest articles, whitepapers and webinars on process mining
- Standardize
Make sure you have an optimal data flow by moving common input elements to feature earlier in the process. Standardize the dataflow and variations so that there is a clear, logical progression for the bots to follow.
- Optimize
Design the process to optimize the interaction between human and machine. Focus on dataflow, input/output - think about how it will work in the future rather than how it happens now. This is your opportunity to build the operational model as you would ideally want it to be, so make sure you have a clear vision of what you want and be wary of making compromises that don't deliver on this vision.
READ MORE: Guide to Target Operating Model design
- Configure
Utilize your existing technology platforms to make the process as lean and efficient as possible. Configure your applications to support a robust (automated) process, involving as much expertise as you can muster.
- Automate
The previous four steps have been a scaffold to build automation over. You are now ready to introduce automation tools such as RPA and remove the remaining manual activities. As you structure your automation, make choices that minimize maintenance and maximize reusability.
- Add AI
In this final stage, expand the scope for automation by adding AI capabilities to your bots. Make sure to have sufficient data to train and test your solution.
Taking time to get these steps right will pay dividends once the automation is under way: laying the groundwork for automation may be slow, frustrating and sometimes counter-intuitive before automation begins. But remember, you are making the processes work better for robot workers, not humans; proper preparation will save you a lot of time and effort further down the line and may well be the difference between your automation failing and being the success you need it to be.
