Robotic Automation: What is it and why should I care?

RESEARCH REPORT // FREE
 
Robotic Automation is a business and technology practice that uses software ‘robots’ to act as synthetic application users to automate highly repeatable, highly structured tasks across business systems. Is it a good fit for your organisation?

Top takeaways

At its heart, Robotic Automation is nothing new (technically)

The fundamental technology that powers Robotic Automation offerings, which provides the ability to capture the structures of applications’ user interfaces and then drive those interfaces through automated actions, has a heritage stretching back to the mid-1990s and automated GUI testing frameworks. However, it’s been extended and packaged in new ways to serve new business operations automation use cases.

Nevertheless, its time has come

Robotic Automation technology has become particularly valued by service providers today because of how it solves ‘legacy’ problems that their clients have typically created for themselves – in a relatively low-risk way. Now, adoption of Robotic Automation is rolling out in a second wave across enterprises themselves – particularly in contact centres, shared-service centres and other back-office administration environments.

Tackling ‘long tail’ integration and automation challenges

Most organisations of significant size struggle to meet demand for new technology, or technology change. Increasingly, business teams are taking responsibility for procuring platforms that can make them more productive, and then self-serving. Where central IT provision can tackle the largest, most demanding problems with centrally-funded platforms, “long tail” problems may never get enough priority to be addressed.

Robotic Automation technology offerings are relatively simple (from a technology point of view) to implement and inexpensive, and offer a compelling way for business teams to solve automation challenges that may otherwise remain unaddressed.

What is Robotic Automation?

The never-ending challenge of legacy systems

Legacy systems have been with us for decades. In the 1990s, mainframe and midrange computer systems were the chief sources of challenge for IT departments as new waves of investment focused on Unix and Windows-based client-server systems; now, the Windows-based desktop systems of the 1990s have joined the ranks of the mainframe and midrange systems. None of them have in fact gone away, despite the best efforts of an industry constantly highlighting their deficiencies. Windows XP, for example, is frequently estimated to retain around 8-10% share of installed operating systems; and almost all of the world’s largest 100 banks still use IBM mainframe computers for core banking transaction processing.

Clearly, these systems are not going away. Partly, this is about cost and benefit; in many organisations, although there’s widespread understanding that these systems would not in ideal circumstances be used, there’s never enough of a business case to make the investment needed to upgrade or rebuild the systems or find a replacement (and also then swallow the inevitable organisational and process reengineering that would be required to accommodate new systems). Partly, it’s also about business entropy; whether it’s because of mergers and acquisitions, or through the never-ending march of tactical business software applications to support new product lines, new promotional strategies or new communications channels, the supply of legacy systems for many large organisations in particular seems inexhaustible.

Although the business case for replacing these systems is often difficult to make – particularly in the context of a wider slate of projects competing for limited IT budgets – the consequences of ever-growing systems estates can be very challenging; particularly for people responsible for administering back-office operations or staffing call centres. It’s not uncommon to find, for example, call centre staff needing to re-key information across 7 or 8 different systems to deal with standard customer enquiries.

Automating the factories of the information age

As organisations in developed markets first began to adopt Business Process Outsourcing (BPO) practices in the mid-1990s, the value proposition of the BPO providers was largely based on labour arbitrage – the simple idea that offshore labour was dramatically cheaper than onshore labour. As use of offshore service providers and BPOs has become mainstream, though, competition for talent among service providers has begun to erode the offshore price advantage – and what’s more, that price advantage can only provide benefits to outsourcing clients once, in the first year of a contract.

Not surprisingly, then, offshore service providers and BPOs have sought ways to do more to lower their cost of operations – and work automation is one obvious way to do this. With many clients’ business processes being taken on by BPOs ‘as is’ – legacy technology investments included – and with little opportunity to invest large sums to re-engineer those processes or redevelop or replace core systems, providers have sought comparatively low-cost ways to automate the most repeatable, highly-structured (or ‘rules-driven’) tasks that occur within the processes they manage. The nature of a lot of outsourced processes means that this kind of highly routine work is concentrated, aggregated and executed at high volumes by service providers, so with the right practice and mindset, carefully-guided automation can make a big difference to operations cost and quality.

This is where Robotic Automation found its first major advocates.

Defining Robotic Automation

We define Robotic Automation as follows.

In the context of business administration processes (as opposed to manufacturing processes), Robotic Automation is a business and technology practice that uses software ‘robots’ to act as synthetic application users to automate highly repeatable, highly structured tasks across business systems.

With Robotic Automation, software robots gather and update information in other business software applications by automating actions against their existing (Windows-based, web-based, or other) GUIs rather than via specialised integration interfaces or hooking into underlying databases or program structures.

Robotic Automation today is most commonly associated with the automation of clerical tasks – data entry, data checking, and so on.

There are two variants of the term Robotic Automation in common use: Robotic Process Automation (RPA) and Robotic Desktop Automation (RDA). These terms describe two types of technology offering that utilise the same core technology, but apply it in different ways. However, there’s more to this space than just RPA and RDA.

Three dimensions of Robotic Automation

The fundamental technology that powers Robotic Automation offerings, which provides the ability to capture the structures of applications’ user interfaces and then drive those interfaces through automated actions, has quite a long heritage. As the Windows client-server application revolution spread in the 1990s, the first GUI testing toolkits began to appear, to help Windows application developers; tools like Mercury Interactive’s (now discontinued) WinRunner pioneered this space. Today’s Robotic Automation offerings build on this heritage.

What is newer, though, is the packaging of core automation technology with design and management tools that are optimised for business operations use cases. This means providing design tools to specify how a robot should automate actions across multiple applications in series or parallel, and providing management tools for robots around auditing, creation of operational reports and dashboards, scheduling and so on.

Different offerings package together these tools in quite different ways (and the differences between RPA and RDA are a part of that variation). As you can see from the figure below, though, today’s Robotic Automation offerings provide capabilities that range in various ways across three different technology dimensions. The sections below explore each of these.

or_what_is_ra_1216_fig1Source: MWD Advisors

The design approach: recording – modelling

Some Robotic Automation design tools are based on interaction recording, where a user records a series of interactions with one or more applications as those actions are taken ‘live’, saves the recording, and then ‘plays back’ that recording to drive automation. Other Robotic Automation design tools are based on modelling, where a user specifies the interactions a robot should instigate ‘offline’, using some kind of design tool. Usually this design tool makes use of flowchart concepts, but not always (some tools require designers to use a scripting language). The resulting models are saved and then used to drive later automations.

The principal advantages of a recording-based approach: it’s easy and quick to create automations.

The principal advantages of a modelling-based approach: it’s possible to specify additional actions as part of an automation – for example, actions that transform data so it can be copied from one application to another (needing data in a different format); or actions that can help with error handling. It’s also easier for teams to share knowledge of how automations are constructed.

The execution approach: interactive – headless

Some Robotic Automation offerings are built around robots that automate interactions with users’ desktop applications by running within each user’s desktop environment itself – working ‘side by side’ with each user. Other Robotic Automation offerings’ robots execute on a separate server (or multiple servers), automating interactions with applications ‘behind the scenes’ and carrying out their work when triggered by other software systems.

The principal advantages of interactive execution: the user is in control, deciding when to use automation, being able to see when automations might benefit from changes or improvements, and being able to intervene if problems occur.

The principal advantages of headless execution: robot-driven automations are used to completely remove certain tasks from users’ concerns, significantly streamlining their work environments and the business processes they participate in. Centralised operation also makes monitoring and auditing data easier to gather.

The operating model: personal – enterprise

Some Robotic Automation offerings are designed primarily to support individual users’ own personal work productivity (or the productivity of small teams). Other Robotic Automation offerings are designed principally for deployment at significant scale, supporting many tens or hundreds of use cases and automating work through hundreds of robots working in concert. These enterprise-scale offerings also typically provide facilities that support large-scale reuse of design artefacts, and tools that help teams manage the impacts of business and technology changes on automations.

The principal advantages of personal operation: the technology is simple to deploy to small groups, and comparatively cheap.

The principal advantages of enterprise operation: if you’re serious about the role of automation in business processes at scale, you need the management tools and capabilities that an offering designed for enterprise operation will provide. A personal-scale offering will become unwieldy to manage at scale and managing risks in the face of change will be very difficult.

A word on RPA vs RDA

As you can see from figure 1, Robotic Process Automation (RPA) and Robotic Desktop Automation (RDA) offerings typically have very different ‘footprints’ in terms of how they support particular design approaches, execution approaches and operating models. Whereas RDA offerings typically focus on supporting personal (or small-team) productivity with interactive execution, and typically do this through recording-based design, RPA offerings implement a headless execution architecture, and typically (though not exclusively) implement model-based rather than recording-based design.

Both RPA and RDA have roles to play in improving operations in organisations, but they offer different value in different use cases.

Why is Robotic Automation important?

Established technology, but new business value

The heart of Robotic Automation technology is mature, but it’s become particularly valued by service providers today because of how it solves ‘legacy’ problems that their clients have typically created for themselves in a relatively low-risk way. Now, adoption of Robotic Automation is rolling out in a second wave across enterprises themselves – particularly in contact centres, shared-service centres and other back-office administration environments.

BPOs and other service providers have delivered large-scale proof of the potential value of Robotic Automation technology, and although it’s by no means a silver bullet, enterprises are now keen to reap the benefits shown here.

Cost, quality, compliance

Software robots don’t need to rest, can work 24×7, and don’t lose concentration. It’s not uncommon for Robotic Automation implementations to lead to labour costs reducing 60-70%. Errors can occur, but in well-designed systems error rates on automated tasks can be very low (<0.5%). Lastly, because software robots often follow predetermined scripts or models, and particularly because their activities can be easily monitored, audited and reported on, demonstrating task completion compliance with key business policies and industry regulations is straightforward.

Customer experience

Most large organisations have major programs of work focused on improving customer experience – and of course key agents in the delivery of customer experience are customer services and call-centre teams. Poor request turnaround times (TAT) and poor first-call resolution (FCR) rates are strong contributors to customer experience problems – and so anything that can make it easier for agents to respond more quickly and accurately is going to be a valuable tool in this context. Robotic Automation technology can not only help assist agents in their work; it can also, in some cases, enable some customer requests and transaction types to be handled in an automated way through an online portal or chatbot channel.

Migration

Robotic Automation offering vendors consistently report that as well as automating operational tasks, some clients use their technologies to better facilitate application migration projects – instead of using more traditional data migration or integration middleware tools. By working through user interfaces to extract data from legacy applications, and also working through user interfaces to populate new applications with corresponding data, Robotic Automation-based migration procedures automatically benefit from validation rules built into applications’ user interfaces. What’s more these Robotic Automation-based migration procedures require less forensic analysis and design work to ensure that the appropriate data is extracted from legacy applications in the most appropriate way.

Tackling the long tail of automation and integration projects

Most organisations of significant size struggle to meet demand for new technology, or technology change. Increasingly, business teams are taking responsibility for procuring platforms that can make them more productive, and then self-serving. Where central IT provision can tackle the largest, most demanding problems with centrally-funded platforms, “long tail” problems may never get enough priority to be addressed.

Robotic Automation technology offerings are relatively simple (from a technology point of view) to implement and inexpensive, and offer a compelling way for business teams to solve automation challenges that may otherwise remain unaddressed.

Download PDF version
RESEARCH REPORT // FREE

Leave a Reply

Your email address will not be published. Required fields are marked *