The evolution of platform programmatic automation


As human beings, we naturally seek each other out and build societies around the notion of community and togetherness. But despite our grounding in societal structure and proclivity for human contact, we also seek out machines and automation.

While some might trace the origins of automation back to pre-biblical times and point to our ability to build water-channelling ‘technologies’, most of us agree that automation first happened during the industrial revolution in the mid-1700s. This era saw the harnessing of steam power and mechanisation, which of course paved the way for the internal combustion engine, electrification and so on.

In the current digital age, we think of automation in terms of software and the use of robotic process automation (RPA). But even RPA has roots that pre-date its modern form and function, so where did it come from and how are its roots related to its recent rapid ascendancy and development?

Scraping for basics

Screen scraping technology has and had existed as a form of data capture and ingestion before the rise of RPA. Sometimes referred to as data scraping at the presentation layer, we would normally now be more comfortable talking about screen scraping a graphical user interface, but the two terms mean mostly the same thing. 

Now becoming a subsumed technology in the age of application performance interfaces (APIs) and their ability to create fluid channels for data interchange, screen scraping has historically been used to get information from one (often incompatible) system to another.

Still thought of as a precursor technology to modern RPA, screen scraping has left a certain amount of its DNA in modern automation platforms, but today’s tools provide a far more granular means of ingesting information using optical character recognition (OCR) for greater accuracy and contextual meaning.

Ever smarter workflows

We must pass a respectful nod to the related practice of workflow automation at this point. With origins dating back a century to early ‘time and motion’ studies applied to workplace roles and processes, the concept of workflow automation shares a history with RPA because both functions are designed to speed up manual data entry and process management. 

This brief history of automation must obviously include machine learning (ML) and artificial intelligence (AI) as base elements.

After a brief starring role in the movies during the 1970s and 1980s, AI spent a period of quieter development until its post-millennial popularisation and wider development in the era of cloud computing. Now an integral part of RPA systems enabling both task and process mining and learning, AI’s role in programmatic automation can not be overstated i.e. it understands and enables us to infer, direct and manage.

From the sum of these developments and through a defined and refined approach to creating a new standard in automation, we ultimately came to RPA somewhere around the turn of the century. 

When we talk about RPA today, we have moved the discussion to business-centric platform programmatic automation. Often making use of drag-and-drop simplicity, we can now harness automation for real world use cases to speed and manage workflows in a more intelligent way.

The intersection of RPA and AI

Although we talk about AI being a constituent part of RPA, we gain a significant and real acceleration factor if we consider modern RPA working alongside AI. This is because RPA is really good at streamlining repeatable, definable and measurable rules-based business processes, but exceptions and outliers will inevitably occur. 

When we can apply AI to RPA to understand more about the context of every element of data inside a workflow, then we can perform more complex decision-making and work smarter.

Applying automation enables an organisation to adopt a new approach to streamlining its processes, which in turn enables it to start moving towards scaling those parts of its business that are most profitable and most innovative. This is the emergence of positive business change brought about by automation in and of itself. Or, to put it another way, it is automation-fuelled transformation.

As automation now shifts from being a tool inside the IT stack to becoming a standard for the way an organisation thinks about operating and innovating, we are witnessing the widespread adoption of prescribed programmatic automation across diverse industries. With financial and insurance industry application use cases often cited as among the most prevalent, RPA has massive penetration in retail, services industries, healthcare, manufacturing, transport, utilities and more.

Our RPA future

Looking to the future, we can see RPA becoming an ever more prevalent element of technology stacks at every level. Key areas for development include innovations that harness and integrate an increasing amount of machine vision technology to capture and interpret real-world images from human faces to places and objects.

As we further develop machine learning and grasp a new era of cognitive computing, the bots that we train and deploy for engagement inside RPA systems will extend above their ability to automate simple repetitive processes; they will ultimately shoulder more complex tasks and learn additional abilities without humans being involved through intervention or programming, even in the case of exceptions.

Embracing RPA at the platform programmatic automation level means moving the technology itself upstream, out of IT silos and into the tactical and strategic planning processes that every department strives to execute. Because customer outcomes have shifted from being tactical to transformational, RPA will take a frontline role at a commercial level to bolster existing digital initiatives. 

A key part of this progression is testing automations to ensure they continue to perform reliably and accurately on an ongoing basis. Once again, at a platform level we can now benefit from automated testing to rapidly and thoroughly test automations and the applications they serve. This testing can be applied pre-deployment and in live production to ensure quality, stability, resilience and performance.

The immediate road ahead for many businesses is a willing embrace of cloud-native technologies. That new imperative for IT platform evolution must now also feature an automation-first approach to building digitally-centred products and services that make both employee and customer experiences and lives better.