From RPA to IA – Addressing the Long Tail of IT Requirements or Digital Disruption?

Robotic Automation has been a key enabler in manufacturing for ages but what about all the talk concerning automation for the office? Just a new way for creating macros and patching up legacy issues? Or a real game changer – a core component of your future digital landscape?

I’ve got a relatively long experience base from enterprise IT – both as a consultant and as a member of IT management teams. You don’t often have the luxury to address every user’s requirements. Sometimes that’s not a big problem – perhaps you already are working on something that takes away the root cause for those requirements. Everybody lives happier afterwards.

However, too often those requirements go unaddressed altogether. After your ERP, CRM, SCM or some other system with a three letter acronym is in place, and after various workflow solutions and rule engines have been set up, there will still be business requirements that are not met. That’s the long tail of IT requirements. And there are plenty of those requirements – perhaps even as much as a third of the totality. Just not with enough value individually and too fragmented to be addressed.

And what’s left for the user is self-help. Often it’s just long hours of manual copy-paste operations and endless repeats. Or a workaround –maybe a complex spreadsheet hack or a set of macros on top of a desktop database. Either way it’s usually an unwanted solution. The user did not want to spend time on a workaround to start with – and now needs to maintain it. The IT department is not happy either – a requirement gets addressed by an invisible solution that might break with the next upgrade.

So what about Robotic Process Automation? Just screen scraping and macros, and maybe some OCR and machine vision sparkled on top? Or a new means for clerical automation creating considerable value for owners, employees and customers a like? And perhaps going well beyond the long tail of IT requirements and even creating digital disruption?

Let’s visit the analogy from the first paragraph and talk about industrial robotics for a brief moment. Machine Learning capabilities are already being applied to enhance industrial robots – for example to reduce setup times and to increase flexibility. As such this is not new but the scale of it is different. The industry is developing physical machines that are able to learn relatively complex tasks and to get better at them over time.

This is important: it might be the setup project for that industrial robot requires one to invest three times the cost of the robot itself. And what if your manufacturing flows change or if you want to introduce new products? The time it takes to reconfigure the robot means time lost in production or delays to that product launch. And if one of the robots decides to have a break the whole manufacturing line is probably standing still. Similar concerns obviously apply to robotic automation for the office – it’s a good idea to design it properly to avoid painting yourself into a corner.

As an interesting side note, in our team at Symbio we have experience of combining physical and software-based robotics. Besides software engineering and product development we also have a considerable business in quality assurance. And with a lot of skilled engineers in-house we just had to go for automation.

So we developed a model based test automation framework with a bit of machine vision capabilities and combined that with a physical robot from a Finnish company called Optofidelity. This allows one to work with touch screens of various form factors and to automate test script generation. Add some solid reporting and analytics and the benefits are obvious. Smart combination of different building blocks delivers the value.

So what if you combine RPA like capabilities with a bit of business process orchestration, rule engines and with some more advanced machine learning capabilities? Throw in a conversational AI platform or a digital assistant and perhaps you are golden. If you are an RPA house, maybe you will call it RPA 2.0 or some similar increment. If your offering is more oriented towards Artificial Intelligence, you are probably talking about Cognitive or Intelligent Automation.

Of course enterprise software is also getting better. Not only easier to set up and to re-configure – the suppliers for enterprise suites are already embedding various cognitive capabilities in their offerings. To help automate routine tasks and to support users with insights derived from masses of data already available. Or just to help the casual user figure out how to use the solution or what the goal of the business process is.

It’s more likely that a typical enterprise will have several intelligent solutions to brighten the day for employees and customers. Some solutions might be generic purpose toolboxes while others perhaps focus on specific tasks such as customer support. Yet further solutions will live within the confines of specific software solutions used by the employees – or particular products offered to customers.

Whatever terminology wins at the end the building blocks are already available. And the first real-world applications are already out there. Some of these are not only addressing the long tail but are actually creating new digital business models and changing the game. Better customer experiences combined with unprecedented scalability at the backend. It’s easy to predict the impact will be considerable.

Please share your thoughts – how long it will take for such Intelligent Automation to mature and become commonplace? Will it create more interesting workplaces for all of us? What kind of difficulties lie ahead? And how could we use such technologies so that our goings are less burdening for the environment?