robotic process automation (RPA) 3 intelligent process automation use cases and how they work


What is hyperautomation?

Hyperautomation is a framework and set of advanced technologies for scaling automation in the enterprise; the ultimate goal of hyperautomation is to develop a process for automating enterprise automation.

Advanced technologies used in hyperautomation include the following:

  • Process mining and task mining tools for identifying and prioritizing automation opportunities.
  • Automation development tools for reducing the effort and cost of building automations. They include RPA, no-code/low-code development tools, iPaaS for integrations, and workload automation tools.
  • Business logic tools for making it easier to adapt and reuse automations, including intelligent business process management, decision management and business rules management
  • AI and machine learning tools for extending the capabilities of automations. The range of tools in this area include natural language processing (NLP), optical character recognition, machine vision, virtual agents and chatbots.

The term hyperautomation was coined in 2019 by the IT research and advisory firm Gartner. The concept reflects the insight that robotic process automation (RPA) technology, a relatively new and massively popular approach to automating computer-based processes, is challenging to scale at the enterprise level and limited in the types of automation it can achieve. Hyperautomation provides a framework for the strategic deployment of various automation technologies (including RPA) separately or in tandem and augmented by AI and machine learning.

Hyperautomation implies a studied approach to automation. A hyperautomation practice involves identifying what work to automate, choosing the appropriate automation tools, driving agility through the reuse of the automated processes, and extending their capabilities using various flavors of AI and machine learning. Hyperautomation initiatives are often coordinated through a center of excellence (CoE) that helps to drive automation efforts.

The aim of hyperautomation is not only to save costs, boost productivity and gain efficiencies through automating automation, but also to capitalize on the data collected and generated by digitized processes. Organizations can plumb that data to make better and timelier business decisions.

Hyperautomation technologies
Hyperautomation uses many technologies, from process mining tools to AI.

Why is hyperautomation important?

Hyperautomation provides organizations with a framework for expanding on, integrating and optimizing enterprise automation. It builds on the success of RPA tools and addresses their limitations.

RPA owes its rapid growth, relative to other automation technologies, to its ease of use and intuitive nature. For example, because RPA mirrors how people interact with applications, employees can automate part or all their work by recording how they perform a task. And because bots mirror human actions, the automated work tasks can be measured for speed, accuracy or other metrics used by companies to evaluate employee performance on the same tasks.

Early RPA efforts, however, had a serious drawback for enterprise use: The technology did not scale easily. Only about 13% of enterprises were able to scale early RPA initiatives, according to a 2019 assessment by Gartner. Hyperautomation forces enterprises to think about the types and maturity of the technologies and processes required to scale automation initiatives.

In Gartner's view of hyperautomation, the focus is on how enterprises can build a process for automating the automations. This separates hyperautomation from other automation frameworks that simply focus on improving automation tools or from automation concepts such as digital process automation (DPA), intelligent process automation (IPA) and cognitive automation, which focus on automation itself.

Hyperautomation takes a step back to consider how to accelerate the process of identifying automation opportunities and then automatically generating the appropriate automation artifacts, including bots, scripts or workflows that may use DPA, IPA or cognitive automation components.

A complementary idea to hyperautomation is what Forrester Research calls digital worker analytics, which also focusses on performance and process: e.g., how to track the cost of developing, deploying and managing automations to compare the cost to the value delivered. This analysis is important for prioritizing future automation efforts. Most RPA and enterprise automation vendors are starting to introduce digital worker analytics into their tools.

How does hyperautomation work?

Rather than referring to one single, out-of-the-box technology or tool, hyperautomation centers on adding more intelligence and applying a broader systems-based approach to scaling automation efforts. The approach underscores the importance of striking the right balance between replacing manual efforts with automation and optimizing complex processes to eliminate steps. 

A key question lies in identifying who should be responsible for the automation and how it should be done. Frontline workers are in a better position to identify boring tasks that could be automated. Business process experts are in a better position to identify automation opportunities that are handled by many people.

Gartner has introduced the idea of a digital twin of the organization (DTO). This is a virtual representation of how business processes work. The representation of the process is automatically created and updated using a combination of process mining and task mining. Process mining analyzes enterprise software logs from business management software like CRM and ERP systems to construct a representation of process flows. Task mining uses machine vision software running on each user's desktop to construct a view of processes that span multiple applications.

Process mining and task mining tools can automatically generate a DTO, which enables organizations to visualize how functions, processes and key performance indicators interact to drive value. The DTO can help organizations assess how new automations drive value, enable new opportunities or create new bottlenecks that need to be addressed.

AI and machine learning components enable automations to interact with the world in more ways. For example, OCR allows an automation to process text or numbers from paper or PDF documents. Natural language processing can extract and organize information from the documents, such as identifying which company an invoice is from, what it is for, and automatically capture this data into the accounting system.

A hyperautomation platform can sit directly on top of the technologies companies already have. One gateway to hyperautomation is RPA, and all the leading RPA vendors are adding support for process mining, digital worker analytics and AI integration.

In addition, other types of low-code automation platforms ,including business process management suites (BPMS/intelligent BPMS), integration platform as a service (iPaaS) and low-code development tools, are also adding support for more hyperautomation technology components.

Hyperautomation vs. automation

Traditional approaches to enterprise automation focused on the best way to implement automation within a particular context. These automations were highly specific to a particular piece of software. For example, workload automation uses scripts to automate many highly repetitive processes. BPM tools can automate tasks within the context of a specific workflow.

AI extends traditional automation to take on more tasks, such as using OCR to read documents, natural language processing to understand them, or natural language generation to provide summaries to humans. Hyperautomation makes it easier to infuse AI and machine learning capabilities into automations using pre-built modules delivered via an app store or enterprise repository.

Low-code development tools reduce the expertise required to create automations. Hyperautomation could simplify the development of automation even more using process mining to identify and automatically generate new automation prototypes. Today these automatically generated templates need to be further enhanced by humans to improve quality. However, improvements in hyperautomation will reduce this manual effort.

What are the benefits of hyperautomation?

Top benefits of hyperautomation include the following:

  • lowers the cost of automation
  • improves alignment between IT and business
  • reduces the need for shadow IT, which improves security and governance
  • enhances the adoption of AI and machine learning into business processes
  • improves the ability to measure the impact of digital transformation efforts
  • helps prioritize future automation efforts

As enterprises master hyperautomation, there are many ways this discipline could be used to improve business operations.

In the area of social media and customer retention, a company could use RPA and machine learning to produce reports and pull data from social platforms to determine customer sentiment. It could develop a process for making that information readily available to the marketing team, who could then create real-time, targeted customer campaigns.

If an enterprise launches a product quickly and digital process automation metrics show strong customer demand for it, the product could be rapidly scaled to help the company grow its revenue. Conversely, if advanced analysis shows that the product fails to gain traction among customers, the company could minimize losses by dropping it quickly.

What are the challenges of hyperautomation?

Hyperautomation is a new concept and enterprises are in the early stages of figuring out how to make it work in practice. Some of the biggest challenges include the following:

  • Choosing a CoE strategy for the organization. Some organizations may do better with a more centralized approach, while others will see better results with a federated or distributed approach to managing large-scale initiatives.
  • Tools. There is no silver bullet hyperautomation software. Although leading automation vendors are expanding their hyperautomation capabilities, enterprises will struggle to ensure interoperability and integration across these tools.
  • Security and governance. Hyperautomation initiatives can all benefit from in-depth monitoring and analysis of business processes that span multiple departments, services, and even country boundaries. This can add a host of new security and privacy issues. In addition, enterprises need to develop the appropriate guardrails for vetting the security vulnerabilities of automatically generated apps.
  • Immature metrics. The tools for assessing the cost and potential value of automations are still in their infancy.
  • Manual augmentation required. A Forrester survey found that only about 40% - 60% of the code for automations could be automatically generated using existing tools. A lot of manual effort is still required and needs to be budgeted when building robust automations at scale.
  • Getting human buy-in. Most automation vendors are pushing the narrative that hyperautomation will augment rather than replace humans, but the reality is that automation eliminates jobs once done by humans. Workers need to be convinced that the robots will not take their jobs in order for these efforts to take off. Also, the various monitoring tools used in hyperautomation projects might prompt a backlash from knowledge workers worried about the potential misuse of that data.

Hyperautomation examples and use cases

A hyperautomation initiative typically starts with a specific goal to improve a metric or process. Here are two examples of use cases and how each would proceed.

In the first use case, a finance team might have the goal of processing invoices more rapidly, with less human overhead and fewer mistakes. A project could start by using task mining software to watch over how human accountants receive invoices, what data they capture and what fields they paste into other apps. This could serve as a template for generating a basic bot.

This template might then be passed over to the CoE team who would be tasked with generating a final bot. This could include integrating an OCR engine to improve the ability to read invoices and an NLP engine to interpret the payee or the terms in the invoice. The CoE team would also oversee quality monitoring initially, followed by an assessment of how much it cost to build the bot and how much it saved. This data could help prioritize other automation opportunities.

Another use case might involve using process mining software to identify ways to reduce order fulfillment times. This would start by analyzing ERP and CRM data logs to identify why some orders are fulfilled in, for example, four hours, while others take four days owing to various exceptions. Process analytics might identify ways of changing the process that would reduce these delays, such as adjusting credit check requirements for established customers. It may also identify ways to automate some manual processes that cause delays in other orders. Once these automations are implemented, the automation CoE team could calculate the total cost of implementing these improvements and track the total savings over time.

Hyperautomation vendors

There are no vendors that currently offer comprehensive hyperautomation technology. However, various automation vendors are expanding their portfolio tools to support a wider breadth of hyperautomation capabilities.

Vendors expanding their automation repertoires include the following:

  • UiPath bought Process Gold and StepShot to build up its process mining capabilities.
  • Automation Anywhere has been developing its own process mining and task mining capabilities for automatically generating bots.
  • Blue Prism has been developing its own internal process mining capabilities and has announced a partnership with Celonis.
  • Celonis, the leading process mining vendor, recently bought Integromat to expand its automation capabilities.
  • Microsoft has been gradually expanding out its hyperautomation capabilities with its Power Automate line of RPA tools and Process Advisor for process mining.
  • Kryon was one of the first intelligent automation tool vendors to incorporate process discovery directly into its tools.
  • ABBYY has been a leading OCR vendor, and has gradually expanded its portfolio of tools to support a variety of intelligent process automation capabilities. It recently introduced various process mining capabilities to expand its hyperautomation tooling.
This was last updated in April 2021

Continue Reading About hyperautomation

Dig Deeper on Digital transformation

Cloud Computing
Mobile Computing
Data Center
Sustainability and ESG