The last couple of years has seen a tremendous rise in the use of Artificial Intelligence and automation software (also called “digital workers” or “bots”). The former mimics human cognition and perception in narrowly defined areas, whereas the latter relieves human workers from repetitive tasks. Every person reading this article has likely interacted with a chatbot somewhere online, and this kind of automated agent incorporates aspects of both categories.
Together, these two categories of software form the area of Robotic Process Automation (RPA). Products in this area have been developed by many significant players in the market, including Microsoft, IBM, SAP, Appian, Automation Anywhere, and UiPath. Gartner has published a comparison of different platforms, and you can find the 2022 update here.
There are many fascinating technical aspects of RPA, and many research papers are devoted to those. However, it’s at least equally worth discussing the societal and business impacts of the widespread adoption of these technologies.
The advent of the robotic workforce
When discussing RPA, “robotic” usually refers to automation software, not physical machines. However, there are some similarities between the two ideas. For example, replacing a data-entry clerk with a document scanner employing text recognition AI has similar advantages and challenges to replacing a warehouse worker with a robot. Both technical solutions are likely to be expensive in the short term, as most cases will require some sort of customization of available products. However, in the long term, both are likely to result in faster processing, lower costs, and fewer mistakes, at least until someone sends in a non-standard form or places the packages on the wrong shelf. This leads us to an important limitation of RPA, namely narrow definition stability and automated processes.
It should be immediately clarified that, by narrow definition, we don’t mean the scope of applications of RPA, as this is widening rapidly with the introduction of new AI models. The limitation here is in the ability of any implementation to adapt to changes. For example, a human clerk happily adjusts some types of non-standard documents to process, whereas a robotic AI will likely reject these to be manually processed anyway. Similarly, for the stability aspect, if the company’s Marketing Department has a habit of regularly changing office stationery design, this will present a mild annoyance to the clerk but may require regular adjustments to the AI by specialists. Camunda, the provider of orchestration solutions, stated, “RPA is a highly maintenance-intensive technology” in a cautionary article.
Furthermore, processes run by automation software will see no further innovation unless there is a human agent involved. As of yet, AI is not good enough to detect that the programmed processes could be improved. Without a proper, long-term strategy and process mapping, there is a risk of being left with some arcane, outdated automation, which nobody understands anymore.
Nevertheless, with careful planning, and proper mapping of the business processes, many companies adopting RPA have found ample cases for successful RPA implementation, with very appealing ROI. According to one study by ElectroNeek, an automation platform provider for MSP and IT companies, ROI from automation can reach 300%, ranging from 50% to 200% in the first year.
What about the impact of RPA on the workforce? According to a survey commissioned by UiPath, and conducted by Forrester Consulting, “66% said RPA restructures existing work, enabling employees to have more human interactions, and 60% said RPA helps employees focus on more meaningful, strategic tasks.” Furthermore:
- The main benefits include increased efficiency (86%), deeper insights into customers (67%), and improved customer service (57%).
- For nearly 70% of organizations, RPA enables employees to have more human interactions.
- More than 50% of surveyed firms report improvement in employee engagement as a benefit of RPA.
- Change Management was indicated as a significant challenge by 82% of respondents, especially regarding cultural issues.
The specter of irrelevancy
The contrast between the two final points in the list above is interesting. On the one hand, employees resist RPA, likely associating automation with layoffs. On the other hand, successful implementation of RPA results in a more engaged and productive workforce. Such discrepancy points to a significant opportunity for education and some PR work for automation before any implementation is attempted. Secondly, it underlines the importance of careful planning, such that any automation is done for the long-term benefit of the organization, not just for illusory short-term gains.
It’s worth further contrasting this with another study by UiPath, which showed that 68% of global workers expected automation to make them more productive. This clearly shows that people are eager to automate the most mundane tasks, open to innovation, but worried about the relevance of their skills in the modern workplace. With the correct strategy, companies have implemented RPA programs that turn their workers into innovators. Looking across available case studies, there are some the common features of such strategies:
- Proper business process mapping across the enterprise enables management to select appropriate areas for automation, taking into account process stability, availability of tools, reliance on human interaction, and the number of special cases. In addition, focusing on good candidate processes builds expertise in the company with easier cases before tackling complex areas.
- The affected workforce"s prior engagement and upskilling enable them to become automation drivers rather than recipients. In addition, having your staff automate their processes with desktop automation tools ensures that you can start small and see immediate benefits.
- The coherent and directed implementation ensures that certain standard approaches are maintained across the organization. This can be achieved, for example, by providing consistent templates for common use cases or by training automation champions, who will apply uniform guidelines in their respective areas.
- Follow-up monitoring of appropriate KPIs allows management to effectively support the implementation and apply corrective measures if required. Therefore, it’s essential to monitor not only the direct automation metrics, like the number of implemented processes and AI effectiveness but also the softer measures, like employee engagement and changes in attrition rate. These latter indicators may help spot problem cases where automation may have a negative long-term impact by driving out human expertise in favor of short-term cost-saving.
At Cognition Shared Solutions LLC, we have worked with several clients who wanted to embark on a digital transformation and automation journey. In each case, we have helped our clients prepare the strategy by mapping out existing processes using our Trilayer Business Process Analysis™ and highlighting areas suitable for large-scale automation and those where caution was needed due to the presence of any of the constraints mentioned in this article. We helped our clients build roadmaps for implementation, ensuring that the four elements listed above were taken into account.
Clearly, Robotic Process Automation is an area to watch. Further developments in AI, Machine Learning, and solutions available in the market will make these technologies key for business success in the future. However, like any tool, they need to be applied with skill and caution to drive the most benefit for the organization as a whole.
If you would like to discuss the opportunities presented by Robotic Process Automation or are thinking of preparing a plan for implementation in your organization, please contact us, and we would be happy to discuss how we can help you in this journey.