Before diving into the SWOT analysis, here’s a short RPA overview as we see it for those that may be unfamiliar with this genre of software. Robotic Process Automation falls under the Business Process Automation category of commercial off-the-shelf software aimed at SME and Enterprise level businesses. These systems typically include advanced and easy to configure data and application integration capabilities along with a workflow design tool. The concept and proposition is relatively easy to understand; users configure RPA to thread together disparate applications and data to minimise or even eliminate time consuming user interactions with and inputs into that process; this represents the “automation” part. While RPA assisted processes mostly rely pre-existing information from applications or other data-sources, inevitably users may have to provide some of the missing parameters. The principle RPA benefit is that as “robots” shield users from mundane, repetitive tasks such as re-keying the same data across multiple applications, allowing them to focus on higher value added-tasks such as decision-making and problem solving. As always, there are catchy industry buzzwords for this approach such as “swivel-chair automation” and “cut and paste application integration” – the bottom line is that for these tasks, robots are significantly faster, more consistent, more reliable and more accurate than humans who can focus on more cognitive parts of the business process.

Strengths: The focus on integration of disparate applications at the user interface (UI) is a potential strength of the approach; the rationale for this assertion is this is how these applications are designed, i.e. to be driven by human beings. Like screen-scraping solutions that predate RPA, integration at the UI level should be an effective mechanism for acquiring information without the hassles of API-level development. Similarly, the keyboard macro solutions of old were effective in driving data for keyboard input. It’s a proven approach. Some RPA systems have “wizard” or “recorder” capabilities that lend themselves to end-user development and configuration. This contrasts with other RPA packages that are aimed at low-level application integration, the principle seems to be that low-level integration affords a higher degree of seamlessness. Another strength of the proposition is that RPA software can also run in the background and these are referred to as unattended tasks where RPA fits with some of the diverse requirements previously handled by middleware or BPM software. Finally, the simplicity of approach is a benefit although as always, complex requirements may spawn complex RPA solutions.

The opportunities presented by RPA appear to be endless. The fact is that many contemporary business processes are addressed using discrete, off-the-shelf, generic applications and tools such as email, CRM, spreadsheets and ERP. Unsurprisingly, the quality and efficacy of data exchange and seamless integration options between these tools and applications varies considerably. As a result, it’s often left to the end-user to manually transcribe (key) data between these systems and tools. Increasingly, there’s a need to incorporate information from external sources or data from thin-client or browser-based applications that may not include an API. There’s also frustration at the many process gaps in enterprise level line-of-business applications as well as the data that’s apparently “locked up” in these systems.

A key area of opportunity is to reduce the load in ensuring corporate and IT governance; there’s a limit to the level of additional recording and cross-checking that can reasonably be asked of human resources executing business processes. Similarly, IT Groups are typically running at close to maximum capacity and so their bandwidth for delivery of enhanced audit and verification is likely to be limited.

With sharp focus on improving user productivity, any solution that can streamline these disparate elements of a business process and provide an efficient, reliable and low-risk way of increasing productivity is likely to be well received by business sponsors.

Weaknesses. So far, so good and so what’s not to like? Regrettably and despite the hype, some RPA tools fall short as comprehensive panacea for business process automation and integration challenges. Specifically, some applications prove quite problematic to access and control even using mature RPA technology. For one reason or another these disparate applications and tools don’t play well together. Faced with such a scenario, the standard approach for RPA vendors is to bring on the developers and tackle the problem at a low-level which, in my opinion, somewhat defeats the point of a rapid solution development message that many vendors are keen to promote.

I’ve observed a second area of weakness: expanding initial, small, pilot RPA projects to a larger user cohort or wider set of processes. Many RPA systems exclusively use Windows desktop environments as a ubiquitous and apparently well-supported environment to stage process automation. After all, it’s reasonable to assert that the vast majority of applications needing automation and integration are Windows based, correct? This association and thinking may not stand scrutiny. While it is the case that the predominant User Interface for end-user applications is MS Windows, it is not the case that all processes depend on this environment as a prerequisite; this especially the case in the context of Robotic Automation as we need as many processes as possible to run on Enterprise-class servers for performance, cost and ease of management.

So, combining these two factors, MS Windows as a prerequisite environment combined with larger sets of users and processes we encounter two of the largest challenges with many contemporary RPA products in an Enterprise setting: scalability and cost. Scalability in the sense that robotic processes typically need to execute on Windows desktops and some systems require one workstation (virtual or physical) per robot; even where robot execution “system pooling” is supported, the nature of MS Windows desktop environments means that it’s typically one robot executing per desktop machine. Again, these can be virtual or physical. This is also an issue at the low end too; to support even a small set of simple robot processes can mean redundant desktop infrastructure needs to be put in place.

Addressing these weaknesses is achievable as long as they’re factored into initial project planning.

Threats: As a relative new concept, RPA continues to be heavily promoted and many organisations have bought into the dream. Not unsurprisingly, Systems Integration companies have embraced the solution as being entirely synergistic and complementary to their Business Process Improvement strategy and business models. However, somewhat predictability, expectations can be pitched (and / or received) at too high a level; the underlying rationale for RPA is to deliver automation to those manual parts of a process that can be easily replicated by computer, i.e. re-keying of data and triggering specific tasks automatically. As noted above, here’s some anecdotal evidence that there’s more to automation that initially meets the eye and that in the pursuit of automation of existing processes, some simple process re-design work may prove more effective.

Justifying the Investment. There’s no doubt that with unlimited resources and time, traditional methods such as web services integration or a BPM to re-engineer the solution would be more elegant. However, limited resource and time are the fact of life in business and so RPA solutions can play a critical role in providing business agility and flexibility. There’s no doubt that if the correct products are selected, an RPA based solution can be as just as robust, reliable and scalable as other integration and automation solutions.

Increased compliance, maintaining high process integrity and ensuring that systems are robust in the face of increasing fraud and misuse will typically yield dividends by reducing the cost of compliance and the any actual breach. Damage to brand or reputation is less easy to quantify but nonetheless, any improvements in this area are typically welcomed by the members of the board.

In the context of aggregate time saving and reduction in errors combined with soft by-product benefits such as flexibility and better process intelligence, the investment in RPA projects can be easily justified.

The real question is not “does RPA deliver” but “can RPA deliver”. Framed this way, the answer is almost certainly yes. However, as with any requirement seeking a technological solution, it’s imperative to fully understand the nuances of the project and to ensure that any technological cul-de-sacs are avoided at all cost.

Embrace-Digital employs a proven business process discovery methodology that will help your business focus on the critical issues around RPA deployment and adoption. Our consultants take a holistic view to applications, systems, processes and people to pinpoint areas for process improvement while taking the time to ensure that projects don’t get out of hand and impose an unforeseen burden on the project.

For queries or comments, please contact Chris Thomas:

christhomas@embrace-digital.co.uk

www.embrace-digital.co.uk | 0333 577 2629

Related Stories.

Return To Blogs