The manufacturing industry is being remade by technology, and firmly at the heart of that transformation is robotic process automation (RPA). Unfortunately, while the desire to improve organizational processes and accelerate digital transformation is continuing to drive widespread adoption of process-discovery tools among manufacturers, many manufacturers are encountering a mismatch between the expected benefits/challenges and those that are actually being realized.
Consider our recent Blueprint Software Systems survey of company leaders in the US and UK, which indicates that while eliminating waste and delivering more task automation were the most common expected benefits of process discovery, they were not the clear front-runners that adopters had projected. Instead, an increase in the quality and performance of task execution, enhanced compliance, and improved task-visibility were cited as the most common benefits manufacturing companies actually realized.
One possible explanation for this mismatch is the amount of time required for manufacturers to manage and monitor the volume of data produced by machine-based task mining, resulting in a significantly longer time-to-value for improved task efficiency and automation. The overwhelming quantity of data generated by these tools might also explain why many companies are turning to more intuitive task-capture solutions for a faster time-to-value.
Discrepancies, however, don’t stop there. The mismatch also extends to the challenges adopters anticipate prior to implementing task-mining tools and those they experience following deployment. The survey shows that more than three out of 10 companies deploying task-mining tools experience security and privacy issues, while only one in five had these concerns prior to implementation.
Similarly, just 23% of respondents anticipated challenges with the time required to manage and monitor task-discovery tools when considering their use; that number rose to 30% after those tools were installed.
These mismatches can be attributed, partially, to the tremendous volume of data the tools produce. The rise in concerns about privacy and security, on the other hand, might best be explained by the way in which task-mining tools work. These tools typically use recorders that sit at the desktop level and record each step (such as mouse clicks, hotkeys, and keyboard interactions) in the various tasks employees undertake. Workers often complain that this kind of approach feels a little too much like “Big Brother” and ultimately hampers their work habits. Departments that are subject to a high level of regulation, meanwhile, have been frustrated by the consent and privacy issues these tools present.
All of this suggests that a better approach for manufacturers might be using task-capture solutions that are human-driven. In fact, the entire process-discovery effort for most companies should start by identifying who in the organization already understands the high-level processes in play. More than likely, these process owners can provide the insights needed to identify—and map out—the higher-level processes that the company currently uses.
By taking this human-driven step, manufacturers can jump-start a structured approach to process discovery and avoid the time-consuming (and expensive) process of using task-mining tools to produce mountains of data, which then must be examined and interpreted to understand how the company already operates. With these higher-level processes identified, manufacturers can move more quickly to capture the low-level details of the various tasks within each process.
By capturing this low-level, detailed task information—which includes all of the actions, parameters, screenshots, inputs, value metrics, and applications with which the task interacts—manufacturers get a better understanding of the services being used, the specific activities being undertaken, and the specific boundaries of those activities. This, in turn, enables them to accurately assess whether a specific task is too complex for automation or a viable candidate that can be converted into a bot for increased efficiency and quality-of-task execution. It also speeds automation development because it is precisely those low-level details that need to be coded in any automation.
Here is where technology comes in. Task-capture tools provide a more cost-effective, non-invasive alternative to task-mining software, enabling manufacturers to identify the low-level details of the tasks that comprise each process, then map the information gathered via a process editor. From there, the company’s developers can further modify and optimize each task, evaluating whether it represents a viable candidate for automation.
Task-capture tools also offer the easiest way to identify the final step in the process-discovery effort: identifying the process’ critical dependencies, loosely defined as all of the applications, business rules, regulatory and compliance constraints, and security protocols that are connected to a company’s processes and tasks. Unlike high-level processes and low-level details, which provide the “what” of any process, critical dependencies are needed to understand the “why” of each process step.
Critical dependencies provide the context needed to understand what is currently occurring and why tasks are executed in a certain way. This, in turn, dramatically improves the quality of both automation initiatives and change-management efforts. If a particular regulation changes at some point in the future, knowledge of critical dependencies enables immediate identification of the automated process and process steps affected by that regulation so that they can be reviewed for compliance.
By taking this structured approach to process discovery, manufacturers ultimately will be in a much better position to align their expectations with reality and speed their digital transformations.
Tony Higgins is the chief product officer at Blueprint Software Systems