Automation, the use of machines to perform work, today most commonly refers to the use of computer technologies to perform the tasks humans would otherwise do as part of their jobs.
The use of computer-based process automation is widespread, with organizations deploying a broad range of software tools to help them reach the automation goals they set as part of their larger digital transformation objectives.
According to the December 2020 Global Intelligent Automation Study from Deloitte, 73% of organizations worldwide use automation technologies. That's a significant increase from the 58% of organizations using such technologies in 2019.
Some of the most commonly used options are robotic process automation (RPA), business process automation (BPA) and digital process automation (DPA). Each of the three technologies offers benefits, and each has distinctions that separate it from the others.
This article is part of
Let's dive into RPA vs. BPA vs. DPA.
What is robotic process automation (RPA)?
RPA technology mimics the way humans interact with software via a UI to perform high-volume, repeatable tasks.
RPA technology creates software programs, or bots, that can log in to applications, enter data, calculate and complete tasks, and copy data between applications or workflows as required.
"Bots communicate with business systems to streamline processes and reduce the burden on humans," explained Anurag Saxena, partner for automation and cognitive at technology advisory ISG. But RPA doesn't inherently have intelligence or decision-making capability, she added, unless it is combined with artificial intelligence and machine learning.
The work best suited to RPA is rules-based. These are discrete tasks done the same way over and over, with no deviations that require human decision-making.
RPA's primary benefits are increased efficiency, lower costs and reduced errors. RPA bots can perform tasks faster and with consistent accuracy and reliability. They can work round-the-clock without taking breaks.
Another reason for RPA's growing popularity in the enterprise is its relative ease of use. RPA works with an organization's existing infrastructure and applications. Also, because many vendors offer low-code/no-code RPA platforms that require little to no programming experience, business users can harness RPA, creating their own bots with minimal help from their IT departments. As such, business users are driving much of the RPA adoption.
RPA, in distinction to other automation technologies discussed here, is also limited in the number of tasks it can reliably perform. Craig Le Clair, vice president and principal analyst at Forrester Research and author of Invisible Robots in the Quiet of the Night: How AI and Automation Will Restructure the Workforce, has cautioned enterprises to observe the "rule of five" in building RPA applications because they tend to break when a bot must make more than five decisions, manipulate more than five apps or make more than 500 clicks.
What is digital process automation (DPA)?
DPA is software technology used to both automate a process and to optimize the workflow within an automated process.
A big focus of DPA is to improve employee and customer experiences by taking friction out of the workflow. The software is used to create efficiencies and enhance UX in various areas of the enterprise, from IT service requests to onboarding new employees and client intake.
Organizations use DPA to automate a process from its beginning to its end. Typically, DPA is used for processes that are longer and more complex than the tasks that can be effectively handled by RPA. These processes can contain multitudes of decisions that, if using RPA, would create bots that are too long and too difficult to maintain.
"The DPA world is about transforming a process; it's about creating a new process," Le Clair said. "It has more of a broader end-to-end view of a process, and the assumption is that you'll be continuing to improve it over time."
In its March 2019 report, "RPA, DPA, BPM, And DCM Platforms: The Differences You Need To Know," Forrester used the term digital process automation as a replacement term for business process management, the long-established discipline that uses a variety of methods to discover, model, analyze, change and optimize business processes.
In the Forrester schema, DPA is divided into two types: DPA-deep, similar to the BPM approach described above, and DPA-wide, which is closely related to RPA:
- DPA-deep is automation that transforms and improves a business process and, because of the complexity, requires skilled technologists to implement and focus on continuous improvement.
- DPA-wide aims "to extend process design beyond small, highly skilled development groups to business users" and can "be managed by the business and delivered using low-code platforms and Agile methods."
What is business process automation (BPA)?
BPA automates workflows within an organization; as one step in the business process is completed, the BPA software then automatically triggers the next step.
BPA software is used to automate complex, multistep business processes that are usually unique to an organization and are part of the organization's core business functions.
BPA's holistic approach stems from the technology's capability to work across the multiple enterprise applications and systems required to complete a typical business process. Gartner describes BPA as automating the processes that "run the business" versus the tasks that "count the business."
With a BPA approach, organizations often first analyze and improve a business process before automating it, which is different from the mimic-as-is tactic typically used in RPA.
Reworked, optimized processes using BPA remove human hands from the workflow; with human workers no longer involved in the automated process, they're not introducing individual workarounds or unauthorized changes to the workflow.
Consequently, BPA is used by enterprises in their digital transformation efforts for the accuracy, efficiency and reliability it brings to each automated process.
RPA vs. BPA: Differences, similarities and possible pairings
RPA and BPA are both automation technologies aimed at shifting work from humans to computers, but there are some differences to remember.
RPA automates tasks, while BPA automates multistep processes, a difference reflected in their cost. RPA's more limited scope of automation typically costs significantly less than a BPA deployment.
RPA can, in turn, be deployed relatively rapidly to automate tasks without having to rework processes to reap ROI. BPA typically requires analysis and improvements within business processes to gain optimal returns.
RPA is more lightweight, with RPA vendors offering low-code/no-code platforms, which enable business users to create bots to automate parts of their job. On the other hand, BPA requires coding and more complex development, thus falling to the IT department to deploy and manage.
RPA can be integrated within BPA platforms to bring further automation -- and thus efficiencies -- to areas such as the customer relationship management process and enterprise resource planning process. RPA bots, for example, can be deployed to populate vendor forms used for procurement as part of the ERP process, where BPA software connects to, streamlines and automates the workflows within the entire resource planning process.
In short, RPA and BPA work together to help support an enterprise's digital transformation.
RPA vs. DPA: Differences, similarities and possible pairings
One of the key differences between RPA and DPA is the extent of automation enabled by the technology: RPA automates tasks, whereas DPA works with entire workflows and processes.
Another difference: While RPA and BPA take a task out of human hands, DPA provides automation that helps humans better interact with the systems so that the users have a better, faster experience as they do their work.
RPA can also be used in conjunction with DPA deployments, with bots performing the repetitive, time-consuming tasks that exist within the larger processes that the DPA solution is optimizing.
RPA vs. BPA vs. DPA: A terminology divide
A word of caution about hard-and-fast definitions of these approaches. Terminology in the automation field is murky at best and can be marketing-driven. BPA, for example, is used by some experts as an umbrella term for the full range of process automation technologies. However, there are different opinions on that term, as well as others in the automation sphere.
In fact, some use the terms DPA and RPA almost interchangeably, using both terms to refer to the use of scripted automation software to mimic human actions in the execution of rules-based "swivel chair"-type tasks, typically where an individual accesses and processes data from multiple applications.
Meanwhile, others define DPA similar to Forrester but add their own distinctions. Pegasystems, for example, announced on its site that "DPA is not another name for BPM, and it's more than RPA -- DPA is an end-to-end strategy for digital transformation."
Others also define the D in DPA as desktop process automation. To further complicate matters, some vendors use the term desktop automation to apply specifically to software robots that reside within an employee's individual computer where the bots perform specific tasks. Other vendors use robotic desktop automation, or RDA, to describe small-scale RPA for desktop applications.
Process automation technologies often rely on other technologies, including natural language processing, to move the work along; optical character recognition and intelligent character recognition recognize written and printed text and transform it into standardized data that automation systems can use to run through processes.
IPA and emerging process automation technologies
Emerging technologies further expand the list of process automation technologies and the corresponding acronyms that CIOs must sort and parse. One such emerging technology to note is intelligence process automation (IPA).
IPA represents an evolution of RPA where automation is combined with intelligence such as computer vision, machine learning and AI to make the automated process "smart."
McKinsey & Co., in its online piece on IPA, described IPA as "an emerging set of new technologies that combines fundamental process redesign with [RPA] and machine learning." The management consulting firm noted that IPA deployments are "becoming more mainstream in enterprise IT organizations," with most IT transformations today including some form of IPA.