Can Robots Eat Spaghetti?

Here’s a stat automation disciples will find hard to swallow: 70% of the problems with implementing automation relate to performance and scalability*. Most Robotic Process Automation (RPA) robots don’t work fast enough or scale far enough! Reality too often falls short of automation’s promise.

Few in the tech community will be surprised. Many believe automation brings IT complexity not simplification, that it doesn’t save money but creates work and cost; instead of modernising legacy IT, automation adds to it and locks it in. They believe it adds to ‘IT system spaghetti’ and creates bloat instead of fuelling leaner, fitter, agile organisations.

They have a point.

Automation has struggled to scale beyond simple repetitive tasks. Many processes are 80% automatable; it’s the final 20% involving interpretation that require the complicated stuff and add cost and complexity. The problem is one of logic. The more bespoke logic is added to automations, the greater the chances of complicated spaghetti systems. Then there’s the problems of governing and maintaining this bespoke logic – which is creating technical debt.

It is time robots learned to eat spaghetti: to provide consumable bite-sized pieces of logic that expand automations beyond the basics without creating more and more bespoke complexity; to reduce the layers of IT systems rather than proliferate them. And it’s time they did all that at lower cost, so more full process automations are economical.

Here’s a thought:

What if there was a new and simpler approach to automation? What if blocks of consumable logic (discrete tasks) could be built around robots like Lego bricks, snapping together fast and easily? What if instead of being plumbed into the IT system, the building blocks could be integrated onto a virtual platform in the cloud utilising modern API-based coding approaches? And what if more building blocks could be added when they were needed, and immediately dismantled when they were no longer required?

It’s not a pipedream. Digital process on-demand is already here and it’s busting the mythology that has built around automation.

It creates less spaghetti because there are no complex bespoke layers of logic randomly configured into the legacy IT; just repeatable building blocks with consistent integration all hosted on a virtual cloud platform.

Legacy systems aren’t further entrenched- rather they are more rapidly dismantled. Using well engineered, loosely coupled building blocks to connect to legacy IT helps to isolate its various functions. This allows legacy to be more easily replaced in steps over time. At the same time, the more complicated tasks that could never be built into Legacy systems can be added through machine learning and other cognitive technologies.

Technical debt is not an issue. Being in-cloud, delivered as-a-service and consumed on-demand there is little management or maintenance - DevOps looks after that. And because the automations are a series of building blocks they can be broken up when no longer required- one good Lego set can build many different things, using the same blocks.

But what about the economics? Here’s the greatest benefit: digital process on-demand drives the business case that automation has so far struggled to create. By rapidly scaling and reusing investment in each “Lego block”, it creates economies of scale. Marginal cost decreases with every reuse of each Lego block.

It enables fuller and more effective automations at much lower cost. Moreover, because it is digital process on-demand, you only pay for the automations you consume.

The economics of process automation now add up.

It all looks good on paper, but is it proven? Yes. Accenture Public Service has helped one of the largest UK Government Departments to pioneer this automation approach for over two years. They can scale automations at speed, daily hit their SLAs involving huge volumes of cases, and realise significant efficiencies.

When COVID-19 hit, the ability to build, deploy and scale automations at pace and scale proved crucial. A repository of “Lego block” components on a scalable platform meant a strategic capability to be tactical when it really counted.

Now that’s real food for thought.

To learn more about this topic, listen to our recent webinar that focuses on “Busting the Big Automation Myths.”

* Forrester.