This is a guest blogpost by Jason Kingdon, chairman and CEO, Blue Prism
Robotic Process Automation’s potential to transform the way we work means every technology vendor under the sun is now scrambling for a place at the category table. However, it’s only the most informed, sharp-eyed thinkers that spot the key nuances in technical advancement, capabilities and delivery approaches of RPA vendor technologies.
This rigorous clarity is required because it’s only when the proof of concept ends and the real work of delivering ‘proof of value’ begins, that any limitations emerge. It’s therefore vital to discuss the key differences and advancements between RPA vendor technologies and approaches, which broadly fall into enterprise-grade, strategic automation, and more tactical desktop automation. These are really important considerations for anyone savvy and serious about enterprise-grade RPA and with big expectations of what it should deliver.
Demystifying RPA approaches
It’s becoming easier than ever to select from increasingly homogeneous RPA-branded desktop capabilities. This approach involves desktop robots doing keystrokes, running scripts and activities against other pieces of software. There are also robots relying on recorded process steps to complete tasks across desktop environments.
Desktop robots are essentially macros and orchestrators: where the bot is a script and a series of scripts are orchestrated. This ‘orchestrator’ approach to RPA is about linking ‘citizen developers’ using recorders to complete useful tasks and accelerate an individual’s work completion.
Let’s be clear though, desktop-based automation is a bottom up ‘citizen led,’ tactical initiative for personal use, typically delivering simple tasks such as filing an expense claim or completing forms. When any automation technology is distributed across desktops and used in personal contexts – it may help the individual, but it won’t help the whole organisation transform work collaboratively at any meaningful scale.
Risks are presented when a single human user is given autonomy over each process recording as this obscures a robot’s transparency and hides process steps. Duplicated over time at any scale and this becomes a potential threat as there’s almost zero clarity for compliance and governance purposes. Also, any inevitable coding introduces shadow IT – with unaudited process models that represent unacceptable “back doors”, security flaws, audit failures and technical debt.
RPA enterprise server
At the other end of the spectrum is the RPA ‘enterprise server approach’ with digital workers (software robots) providing full stack, multi-agent, processing capabilities that share ‘human readable’ process description documents with each other. This approach uses centrally managed programmatic interfaces, offering transactional integrity and robust process management across the full enterprise estate. It’s similar to a programmable infrastructure sitting above the ERP systems and process management infrastructure. Enterprise applications are treated as passive data services by digital workers that they dynamically unify to deliver full business process automation across the estate.
Unlike desktop bots, digital workers aren’t scripts or macros, but an out of the box, increasingly versatile, highly productive, self-organising and multi-tasking work resource.
Trained and run by business users through a centralised, collaborative platform, it’s a workforce operating independently or seamlessly with humans to deliver complex front-to-back-office data-driven work, with unmatched speed and integrity.
Digital workers also work without interruption interoperating with any system, of any age; uniquely using APIs in the same way human workers use UIs – and without requiring any coding or system integration effort. As all technologies are uniquely made interoperable by digital workers; this makes them easy to augment with the latest and greatest intelligent skills. This unique universal connectivity capability also means digital workers are being increasingly used as the gateway for organisations to easily, safely and swiftly, test, advance and deploy any new technologies for work transformations.
This type of intelligent automation facilitates a deeper understanding of the demands of a modern enterprise organization, resulting in the highest levels of RPA performance, security, resilience, usability, interoperability and governance capabilities. It’s an approach that delivers strategic, end-to-end, work process automation, such as mortgage applications, trade finance management and even company IPO regulatory filing.
Enterprise RPA in action
RPA’s value can only be realised by not only enabling enterprise-led, ‘top down’, smarter ways of working, but applying these capabilities centrally, securely, collaboratively, with greatest ease, at greater speed and scale too. Over time, we’ve seen these crucial differentiators consistently proving the difference between those organisations experiencing major digital work transformations – and those individuals experiencing tactical benefits.
These capabilities mean vast time and resource savings are being generated over 2000 of the world’s biggest and brightest. For example, Telefonica is targeting 1 billion euros, with 100 million euros saved in the first year and FDA filings for Pfizer meant $35 million saved.
These organisations and many others are also experiencing greater operational agility, performance and efficiencies. Many are going further and swiftly solving increasingly complex work problems previously deemed impossible due to time cost and resource limitations. Others are creating unlimited high value growth opportunities with a predicable quality that continually exceeds stakeholders’ ever-changing demands. All are gaining major advantages so they stay ahead.
Ultimately, both desktop and server RPA approaches are fresh and offer new models for productivity. Both offer a way of democratising automation for non-technical business users – but address different levels of business issues. Looking forward, the big question is do these models co-exist, or do they converge, or is it robot wars? If the latter, then we’ll need greater clarity on what is a software robot. Gaining these insights will be the key to ensuring that users fully appreciate which flavour of RPA can actually deliver on its promises.