Intelligent Process Automation and Robotic Process Automation Need Business Process Management
There’s a lot of excitement and confusion over innovations in process automation. The latest may be the emerging technology category of Intelligent Process Automation (IPA). IPA refers to the application of the new AI- and machine learning-based technologies to Robotic Process Automation (RPA) to make it smarter and more flexible. But none of these tools are, properly speaking, about process automation. They all help automate tasks — repetitive, routine tasks. For simpler business processes, automating some tasks may be enough to yield an entirely automated process. However, for most complex business processes, you need to do more — you need business process management (BPM). This post explains the difference between IPA/RPA and BPM and where each is most useful.
Technologies Included in IPA
IPA includes such technologies as:
- Robotic process automation (RPA): tools that automate routine human tasks by mimicking the manual steps a human worker would take when using existing application software. Examples include document and data download, data extraction, data entry, data aggregation, and transaction processing — e.g., back office clerk activities in financial services and insurance, call center, and other swivel chair activities.
- Machine learning/advanced analytics: algorithms that identify patterns in structured data, such as daily performance data, through human “supervised” and automated “unsupervised” learning.
- Natural-language generation (NLG): software engines that create interactions between humans and technology by following rules to translate observations from data into prose.
- Cognitive agents: technologies that combine machine learning and natural-language generation to build a virtual workforce of agents capable of executing tasks, communicating, learning from data sets, and making decisions based on sentiment detection.
Fix the Process with BPM
What the IPA tools all have in common is that they are task automation, not process automation. They all do work, but not flow. You get more bang for your buck if you focus on improving processes, which requires a more holistic approach than just focusing on discreet tasks. BPM is this more comprehensive approach. (It’s not the only one, however. In a future post, I’ll detail more process automation technologies ranging from the most simple to the most complex.) BPM solutions are platforms used to develop, deploy, monitor, and optimize multiple types of process automation applications, including processes that involve both systems and people. IPA should be part of your BPM approach, but it doesn’t replace it.
BPM solutions are platforms used to develop, deploy, monitor, and optimize multiple types of process automation applications, including processes that involve both systems and people.
Effective BPM solutions must address a wide range of situations, enabling human interaction and also orchestrating automation, all of which will be tied to business rules and policies. Ultimately, an enterprise-wide BPM solution should accommodate a broad mix of processes, including those that may be one or more of the following: of varying duration, event-driven, metric-intensive, system-to-system, human-to-system, inter-departmental, inter-organizational, integration-focused, and orchestration-focused.
Distinguishing Characteristics of BPM
BPM has several characteristics distinguishing it from IPA/RPA. Some of these differences are a matter of degree — BPM is more scalable than IPA/RPA — while others are a matter of kind. For instance, RPAs don’t do end-to-end transaction management.
- Process application design and development: BPM process design is more complex to develop than RPA task design (though comparable to some of the other IPA tools). Most vendors have focused on “low code” development capabilities.
- End-to-end transaction management: This is the most significant difference between BPM and IPA. It’s about managing transactions from beginning to end, even if initiated from, or involving, other systems and applications. Many more primitive workflow solutions can only manage transactions within their respective “sphere of influence,” but critical functionality for BPM applications is the crossing of LOBs. For example, wealth management involves retail banking, trust services, and others.
- Support for multiple integration services: Many BPM applications require the integration of multiple business systems within one process, such as multi-account processing in financial services.
- Support for data aggregation and transformation: This is the aggregation and transformation of historical and real-time data from multiple systems for monitoring and reporting. It’s important in applications that cross multiple LOBs, such as consolidated customer service applications.
- Support for content management functionality: Many applications (mortgage, claims, and payment processing) require imaging, document management, and records management functionality. In the past (before 2010), the trend was for ECM vendors to bundle BPM and more basic document workflow along with their content services. Now, most large organizations use separate BPM and content systems, which may be as tightly integrated as the earlier single-vendor combinations.
- Business activity management (BAM): These include process monitoring, analytics, and simulation components. IPA suites provide some of these capabilities, but BPM platforms focus on entire processes, including cross-process monitoring.
- Architecture that supports high performance, scalability, and embedded failover/reliability functionality: BPM applications require a high degree of fault tolerance with ability to spread processes and tasks across multiple servers. There should be no loss in user work or transaction integrity with component failure. Expect native load balancing of transactions and work items and automated recovery.
- Template applications with pre-built components: Templates include processes, rules, roles, reports, integration components, client interfaces, and best practices. The most mature templates are for financial services (loan processing, payment processing), insurance (claims processing), manufacturing (procurement), pharmaceutical (clinical trials), and horizontal applications (customer services, AP/AR).
While there is increasing knowledge of and appreciation for IPA and RPA, it’s important to remember that they are pieces of a larger business process management puzzle. Find and use the best-fitting technologies to get better enterprise-wide results.