Original article was published by District Data Labs on Artificial Intelligence on Medium
By Andrew Pearson
Today’s IT systems have become incredibly complex, with distributed architectures, multi-cloud containers, microservices, Big Data, in-memory systems, and real-time streaming all working together to support business operations. It is almost impossible to keep up with it all.
Many of the solutions that monitor these systems are single-focused diagnostic tools that only look at the historical past. They lack predictive and prescriptive analytical elements that enable proactive measures.
Technologies like Application Performance Monitoring (APM) and Business Process Management (BPM) have evolved from early generations of IT management tools to help companies enhance their operations and reduce time-consuming processes. Businesses are always looking for ways to cut costs, and BPM and process mining have facilitated that by providing visibility and insights into IT systems and procedures.
More recently, Robotic Process Automation (RPA) tools have provided companies with the ability to utilize this information and turn it into actionable intelligence that can streamline business processes, thereby increasing productivity and further reducing costs.
The next-generation solution — Hyperautomation — promises to offer real-time intelligence about the organization, which will provide substantial business benefits and potentially drive significant business opportunities. This future is believed to be right around the corner. Not to sound too metaphysical, but Hyperautomation goes beyond noise reduction, error detection, and in-depth root cause analysis, and towards system self-healing. A worthy IT goal if there ever was one.
The Continual Evolution of Automation
As our IT systems have become more complex, so have the underlying process systems that underpin them. The most traditional IT automation technology — macros — have been available for decades. Like Adobe Premier’s macro functionality, which allows users to crop, adjust, and rename photos automatically, these processes can automate repetitive tasks and systemize complex business applications, databases, or even in model-building procedures. Macros act like keyboard shortcuts that eliminate the need for a human operator to carry out the work.
However, macro sequences must be established in advance and are strictly limited to highly specific processes that do not allow much variation. Some additional human intervention is required to kick off the process and ensure quality.
Business process management (BPM) aims to improve business processes by analyzing them, modeling how they work under diverse circumstances, implementing improvements, and monitoring the enhanced processes. Traditionally, BPM includes process workshops and team interviews, but these methods tend to suffer from employees fearing to express dissenting views, offering up counter-arguments, and groupthink.
Process Mining and Automation
Process mining gets around this bias by utilizing specialized data mining algorithms that detect and categorize trends based on event log data. The aim is to gain insight into a business’s processes by looking at the data running through the IT systems. This can help a company better understand what is happening and improve the efficiency of its operations.
Whereas process mining provides a state of the business as it currently exists, process automation (or business process automation) uses digital technology to carry out a procedure, either individually or as part of a workflow. With the expanded use of technology, an increasing number of business activities can be automated to benefit the business in many ways.
The benefits of process automation include:
- Increased productivity
- Reduced operating and labor costs
- Error reduction
- Increased collaboration
- Speeding up and optimizing task executions
Most departments within an organization will be affected by process automation, with production, supply chain, inventory control, administration, sales, marketing, HR, and IT all benefitting substantially from the technology.
Robotic Process Automation (RPA)
RPA could be considered an addition to BPM rather than as a step up. It leverages automation software, which allows users to configure scripts that run in an automated fashion. The resulting “bots” mimic or emulate tasks such as data collection, uploading or downloading data, manipulation, alerting, response triggering, or model building, all within an overall business or IT process. RPA can play an important role in the automation process, but it won’t fully replace BPM as it isn’t an end-to-end solution.
RPA tools operate on a computer system’s user interface just as a human would, turning a computer into a functioning individual who responds to a set of repetitive instructions. RPA aims to replace humans in an “outside-in” way (i.e., taking control of the keyword) rather than in the classical “inside-out” approach, where tasks are executed by code dictating the steps. RPA carries out coded instructions on structured data, either through GUI interactions or via an API.
The main goal of an RPA is to reduce the burden of simple, repetitive tasks on employees, freeing them up to handle the more complicated and creative tasks in which humans thrive.
Automation of IT Processes (ITPA)
Automation of IT processes (ITPA) is a step above RPA. It improves efficiency by reducing the manual work required to measure and execute routine IT tasks, such as updating or patching systems, configuring new servers or network devices, performing backups, or implementing security policies. Automation can be carried across multiple layers of technology. ITPA allows for consistency between different versions of software or systems as well.
Many see RPA and ITPA as almost comparable as they both leverage software tools to automate processes, but this is where the similarity ends. The differences lie in their complexity. RPA is primarily used to automate day-to-day tasks that are typically handled by humans. In contrast, ITPA is used to automate more complex workflows that are overseen by experienced IT professionals. For example, ITPA might help automate incident management by analyzing incoming alerts, verifying and prioritizing them, and notifying the departments that are being affected. Once the problem has been resolved, the ITPA solution will complete the workflow and close the ticket.
Hyperautomation (RPA + AI)
Beyond process automation, process mining, RPA, and ITPA is Hyperautomation, which involves a combination of RPA/ITPA and AI. Listed as Gartner’s number one technology trend for 2020, Hyperautomation combines machine learning with automation software and often refers to not only the breadth of tools but the steps in the automation process itself (discover > analyze > design > automate > measure > monitor > reassess). In other words, hyperautomation deals with the application of advanced technologies to increasingly automate business processes and augment humans.
One of the biggest differentiators between hyperautomation and RPA, ITPA, and the other technologies that preceded it is that hyperautomation requires knowledge work (i.e., thinking, designing, and evaluating). Machine learning might be the technology that underpins it, but hyperautomation requires hyper-agile working practices and tools.
The result of hyperautomation is often a digital twin of the organization — a digital replica that allows a business to visualize how its processes, functions, and KPIs all work together to create organizational value. A digital twin provides real-time intelligence about an organization, offering substantial business understanding and driving significant enhancements in how an organization operates. Hyperautomation utilizes one of AI’s biggest strengths — its ability to take incoming data and improve its processes and predictions in a virtuous cycle.
The Hyperautomation Edge
We’re on the edge of a hyper-driven world, so hyperautomation should fit right in.
With hyperautomation, a call center can use RPA and AI to automate an agent’s mouse-click process to launch an application that pulls up information from multiple source systems to provide a holistic view of the client.
A mortgage leader uses automation and RPA to accelerate its loan origination process, having robots collect information from various sources. The company found that RPA alone was inadequate, and hyperautomation compensated for RPA’s deficiencies. By utilizing ML and machine vision, hyperautomation could extract important customer information from a diverse set of documents that would have been impossible with RPA.
Manual back-office tasks like data entry and document verification are two other processes ripe for hyperautomation, as is data, image, and video extraction.
For the IT department, the only constant is change, and big change is coming. Adding in the seven V’s of Big Data — Volume, Velocity, Variety, Variability, Veracity, Visualization, and Value — it is becoming clear that IT departments now have a massive task ahead of them to keep everything functioning correctly. The benefits of process automation, RPA, and hyperautomation include increased productivity, reduced operating and labor costs, a reduction in errors, increased collaboration, and, above all else, a deep and detailed understanding of the business.
Gartner sees a cost reduction of 30% within four years for companies that can combine hyperautomation with redeveloped operational processes. With the consistent growth of online services, a move into the cloud, an expansion of microservices, real-time streaming technology going big time, Big Data turning into Fast Data, and IoT about to complicate things exponentially, things are about to get a lot more complex in IT.
Hyperautomation will soon separate those who embrace change and those who don’t, with the latter being left far behind. For those with vision, it’ll be full speed ahead, into hyperautomation-drive.