Various prominent technology analysts and publishers speak of Robotic Process Automation a completely new 'thing', 'trend', or 'wave'. The thing is: RPA truly is a methodology breakthrough in Business Process Automation. Let's consider that for a minute.
This RPA breakthrough is not particularly arising from a new core technology, but from a group of collaborating technologies that are emerging in a robust and enterprise-grade set of vendor supplied systems. Yes, think UI Macro Recorders and Screen Scrapers combined with workflow engines. Then magic happens.
In case you wanted to look more closely into the haystack, you might spot other core integrated technologies, for example: Business Process Execution Language (Oasis WSBPEL), and Microsoft Worflow Engine or simply Microsoft Flow. It is truly fascinating to witness the emergence of RPA as a result of industry and academic research innovations.
With RPA, other previously peripheral technologies such as Optical Character Recognition (OCR) and Image Recognition (IR) seem to have found their way to the BPM space. With this, OCR/IR are two examples from the plethora of artificially intelligent functions foreseen at the third generation of RPA. That makes part of the transition to more Intelligent Process Automation.
Naturally, RPA is rooted in Business Process Management as a practice, and positioned as a new -additional- approach to performing business process automation.
So, what was process automation before RPA?
.Well, traditionally speaking, other than most IT systems being automation engines themselves. BPA at that time was kept for the few enterprises that have reached a top-tier maturity in their business process landscape. Simply because achieving process automation required an existing base of streamlined automation-ready business processes, in real terms, that meant having modelled business processes running in a workflow engine. Many have just never reached that level of business process maturity, or did not want to spend the considerable budgets and time associated with entering the BPM/A space.
Historically, the 1990s largely built on the holistic approaches of the 80s like Total Quality Management and therefore brought mammoth-like enterprise BPM systems such as early IBM BPM and similarly early Oracle BPM. Interestingly, the same era has brought some of the most prominent mammoth legacy systems we face today - the ERP. Reality is, that era was notorious for BPM, largely due to long-running BPM initiatives with big budgets and less than expected business benefits.
Moving along to the 2000s, a time of what could be described as model-orchestrated process automation platforms such as Pega BPM, TIBCO BPM, Bizagi, Camunda BPM, Bonita BPM... etc. It may be fair to think that BPM of the 2000s had recovered from some of the notoriety it had gained in the 1990s. Yet BPM growth at that period suggests that some damage was never recovered.
Surviving the 2000s as a 'BPM software' company had to be considered lucky. It was a time where no major technological breakthroughs had occurred for some time. It was also a time taken for agreement and standardisation, considering the amount of BPM notations and standards which have emerged then.
Negative press on BPM was climbing once again to the peak. Business interest in BPM was somewhat eroding. Between 2004 and 2011 the search term Business Process Management had lost 69% of its global search points on Google Trends.
Having that in mind, one may attribute the BPM technology developed at that time to the to the re-transformed domain of Rapid Application Development (RAD) with the likes of Appian, Mendix, Outsystems, Betty Blocks... Particularly, the BPM focus on rapidly and graphically producing process applications with low code, accompanied with the shift from big-bang waterfalls to agile methodologies had fuelled the foundations of new RAD platforms.
This transition period was a true gamble; and somewhat dependant on a fundamental change to the industry landscape.
It’s January 2015 and Robotic Process Automation is practically unheard off. Fast forward through 2016-17 and the RPA market is forecasted to reach $2.9 Billion by 2021 (Forrester), a slightly reserved number in comparison to other sources.
Meanwhile, on the sell side, the largest RPA software providers are entering late rounds of $30-40 Million funding bringing in somewhere from $70-100 Million each. At least 20 considerable new RPA technology providers have emerged in 2016-17, effectively doubling the overall market landscape. Put all the investments made together with the racing RPA providers and think how this cash is put in a market segment which was mostly non-existent a couple of years earlier. Having reached unicorn valuation by early 2018, the story of UiPath reflects the level of success in this industry.
On the buy-side, the prevailing thought of the first large enterprise adopters was mostly like "This RPA thing seems to be taking off. Maybe we should get someone in to pilot with an 'easy process' and see for ourselves". RPA technology and concepts were proven, and there are several successes in enterprise adoption. Yet, this is the time to scale. Some might think to expand their proofs of concepts to more processes, but what does it really take to scale RPA successfully?
Is this a 'just scale it' paradigm?
The Paradigm Shifts
There are two big paradigm shifts in the RPA space. One is introduced by RPA itself, and another stems from the capability of adopters to position and administer RPA in their businesses.
The first is a shift in the outcomes of performing process automation. Previous BPA was effectively producing more interfaces to be maintained by enterprise users (i.e. it was increasing resource intensiveness). On the other hand: RPA is pragmatically automating existing interfaces and taking the burden of maintaining them. With this shift, RPA is suited for automating the iceberg portfolio of legacy applications, and with that, decreasing resource intensiveness.
The second shift was evident during the 2017 obstacles to scaling RPA in businesses. Yes, it was late-2017 and RPA was stumbling (McKinsey). Some suggested an adoption rate of just 4%. Several robotics programs were put on hold, or CIOs have flatly refused to install new bots (McKinsey). This shift is seen in the position of initial adoption; and later in scaled adoption, and subsequently in the operation of scaled RPA in business.
It is useful to consider this shift from the perspective of the simplistic business-IT spectrum. As such, many of the existing BPM practices began with IT and then transitioned to the intersections of business and IT. However, RPA is first starting with the business side of the spectrum. This shift ushers a variant set of challenges at each stage of the adoption. In fact, where this adoption begins or "mushrooms" in the organisation is as important as how to manage it, scale it, and then achieve the enterprise capability of Digital Process Operations.
The Success of Scaled RPA
Many enterprises are already pursuing RPA programmes at organisation scale. Additionally, Free RPA tools and training resources are available, the use of these tools by employees is showing what could be signs of a more natural digital transformation. As employees are looking to free their time from their most daunting ‘robotic’ activities.
Either way, with an RPA programme or with a more natural adoption, the success of RPA at scale depends on achieving the second of the paradigm shifts in a way suited for each business organisation respectively. There isn’t a one size fits all when it comes to building a digital future of a business, but there is a slow and bumpy transformation towards that future.
Positioning RPA in its right place in the enterprise, within a larger BPM practice, as an additional means to automating business processes could be key to this success, and may smooth the bumpy way to achieve the goal of a more digitalized business.
I'd suggest starting with building a Lean RPA Practice or adding Adding RPA to your existing BPM practice.
An earlier version of this article is published on the Process Excellence Network - PEX.