Skip to main content

In a recent blog entitled RPA 2021.5: Its Go Time! I observed that we are at a watershed moment in human history.  For the first time, technology is going to work for human’s vs humans working for technology.  That’s all well and good, but how do we get there? 

Since the early days of automation, RPA, scripting or “screen scraping” has struggled with both reliability and scale.

Because of this, the process of adopting this technology hit roadblocks and resistance..  Even currently, as UiPath and others showcase use case after use case of success, some remain skeptical,or just unaware of the technology itself.  

To be clear, regardless of which RPA vendor a business chooses for a project, most activities used to automate applications utilize the same techniques to interact with windows, web, character based and Citrix applications.  While automation challenges still exist with some applications, they are lessening each day. The use of AI, as an example of an application using advanced Computer Vision and anchoring techniques, has moved the mark, making this technology reliable for those who are hard core RPA Developers as well as Citizen Developers.  As a developer who in its very recent past, struggled to “crack the nut” on some web applications (like Salesforce) and dealing with Citrix and its unreliability, this is all welcome news.  UiPath has stepped ahead of the pack with this technology.  I credit Daniel Dines for pushing this forward.  

So, what does this mean? Quite a bit, actually.

Before, RPA was a back-office tool. Now, most in-production bots performing high value tasks are scheduled.  These generally don’t involve a lot of maintenance or human intervention.  Rules-based Bots have limited abilities because they are entirely rules based.  To be frank, one can build only so many “if, then and else” rules before the logic gets too complicated to maintain. Machine Learning and AI can help with some of these decisions, but in order to take advantage of this, robust datasets must be available.  This in itself no small task.  Many organizations do not have the data to build robust models.  Or, because it’s so fragmented, the data must go through an extensive cleansing process before use. Therefore, I see an increased need for either a “human in the loop” and/or a solid Citizen Development program to make these Bots more intelligent.  

I feel the technology is solid now.  Organizations can create efficiencies by adopting “top down” and “bottom up” strategies to tackle the work backlog.

Citizen Developers can alleviate the strain on IT.  This provides the business with solutions. This bottom-up methodology allows for most simple tasks to be automated by those that understand the workflow the best.  In addition, they can manage and share these bots with other members on their team. The automation team handles overly complex tasks. The team determines whether the task is worth automating and applies the appropriate resources.  The other way of thinking of it is from the top-down where the business has identified a backlog of tasks requiring automation.   The larger team can determine which can be given to the business i.e., Citizen Developers to automate and the rest can be done by the RPA development team. 

The reasons why organizations need to adopt this mindset immediately are growing, and the results will be robust. True Digital transformation cannot occur without dealing with the minutiae first.