Speeding up RPA through behavioral observation
How behavioral observation can help to optimize the RPA journey and enhance efficiencyAdd bookmark
As more businesses leverage robotic process automation (RPA) to address individual legacy system components and the pain points they create, it is worth considering how to best introduce and integrate these tools to your advantage.
At CamundaCon LIVE 2020.2, Andres Jimenez Ramirez, professor at University of Seville and Hajo A. Reijers, professor at Utrecht University, explained how implementing RPA with a human-centered approach will eventually pay dividends.
Their research explores how when automating manual tasks, RPA typically mimics the behavior of human employees. To introduce automation tools, most commercial platforms rely on a manual analysis to identify the tasks suitable for RPA.
However, manual analysis is incredibly time-consuming. Not only does it require documentation analysis, it also necessitates determination of expected case frequency, potential exceptions and various other attributes that are complex to estimate. Therefore, when a company wants to introduce RPA to automate a low-level manual task, a far deeper analysis of the process is required.
Optimizing RPA through human interaction
In order to optimize the RPA journey, a human-centered approach is required. The knowledge that human workers gain over time to review errors and continually improve processes can be transferred to RPA.
RPA is focused on interaction of the user interface (UI) to automate low-level human tasks. Before it can be introduced, businesses need to ‘mine processes’ and automatically discover the real steps in each task as they are executed by a human worker. Automation can also be extended to process mining itself.
Mapping processes with behavioral observation
Behavior observation allows process owners to better understand how human workers tackle the process and exceptions that occur. Source: Camunda
Through behavior observation, process owners can attain an enhanced understanding of how human workers tackle processes and be aware of the exceptions that occur.
One starting point is to log information to capture various attributes of human behavior, by monitoring the interactions between human workers and the information systems within the RPA implementation project.
Even when you do not have access to the source code of the software your current processes are running on, you can still utilize desktop apps to capture information. The application of activity monitoring, screen captures, and mouse and keystrokes to obtain a complete log of all systems is the most critical part of the process, and enables in the generation of a behavior observation log.
The behavior observation log provides comprehensive data that is difficult to interpret. Source: Camunda
While the behavior observation log provides comprehensive data, it is difficult to interpret and it does not give you the ability to identify events with a high level of similarity, thus which events should form part of the RPA implementation. This is where observed behavior, in the form of screen captures, clicks and keystrokes, can be transformed through automation into a meaningful UI event log, or UI log, through image-similarity techniques.
For example, the analysis of captured screenshots is important, but very time-consuming. A way around this is to reduce the image size and colors down to an array of bits for comparison, much like a fingerprint. Then you can begin to separate the events into cases using the ‘Hamming distance’, meaning how many bits are different in one image to another, and identify which cases comprise the individual steps of the complete process.
Because human workers will do other things aside from carrying out the process, such as checking and responding to email, the first process model you return will resemble spaghetti. This necessitates cleaning the log to discover which activities comprise the process. After a period of analysis, you can produce an 'as-is' design for your RPA bots to follow.
Without automating the behavior observation, process mining can take weeks or months. Automated behavior observation allows process stakeholders to save time, discover aspects that might not be easily transferred to an RPA bot, and find new ways to manage these.
Behavior observation log analysis leads to an 'as-is' design for RPA implementation. Source: Camunda
RPA challenges to overcome
The design and development of RPA bots can be accelerated through human behavior, even in very challenging situations, such as with little access to data or when working with screenshots. However, organizations still need to overcome some challenges.
If you observe human workers and they each execute the task in a different way, you will run the risk of automating inefficiencies by not understanding the process itself, due to a messy behavior observation log. There are many exceptions that require humans to take over from bots, and in these situations analytics can be applied to facilitate this and, in turn, assist humans workers with more support and insights.
You can watch the full presentation on speeding up RPA through behavioral observation on-demand here. You can also read more about Camunda’s integration with UI Path and RPA functionality in this blog.
Also watch the on-demand RPA Modernization with Camunda webinar, presented by CEO Jakob Freund, to learn how RPA modernization provides a lifecycle approach to orchestrate, choreograph, analyze and monitor your RPA bots today. It also provides an architectural path that replaces RPA bots with API and micro services for the future.
This article was orignially published as Speeding Up RPA through Behavioural Observation.