THANK YOU FOR SUBSCRIBING
Robotics Process Automation (RPA) - Experiences from Today, looking towards Tomorrow
By JONATHAN KIDD, HEAD OF DIGITAL OPERATIONS & ROBOTICS, BANK OF IRELAND
1) Perceiving (the senses)
2) Processing (the doing)
3) Predicting (the thinking!)
In this model, RPA sits firmly in the Processing stream. More broadly, this perceive-process-predict model allows us to position other capabilities which are considered aspects of AI within an Intelligent Architecture framework.
So, coming back to just RPA. What have we learned about RPA over the years? What are some of the key observations that can help us to make better use of it in the future, either on its own or as part of a broader Intelligent Automation toolkit?
1. A journey to Intelligent Operations
As the diagram illustrates, RPA is one element in a toolkit which facilitates increasingly digitised end-to-end journeys that sit above or alongside traditional enterprise IT. This suite of tools can be categorised by the type of human behaviour they mimic – Perceptive, Process, or Predictive. What success we have had to date with RPA has been built on the availability of a number of these components including digital channels, data transformation, digital workflow, and digital communications. The more components, the more interactions that can be automated; and the more human functions can be mimicked.
2. Which came first, the Robot or the LEAN process?
Should I LEAN before I automate? The risk of not Leaning first is that we obfuscate spaghetti processes even further by embedding them in RPA tools. The risk of Leaning is that we realise the value of RPA too
The risk of not Leaning first is that we obfuscate spaghetti processes even further by embedding them in RPA tools. The risk of Leaning is that we realise the value of RPA too slowly and give up!
slowly and give up! My experience has been that significant benefit can be realised by automating key processes without optimising them; building
the necessary RPA capabilities as you do that. At some point, the evolution to automating end-to-end journeys will become a natural transition as the combination of RPA, Subject Matter Expertise (SME), Agile and LEAN knowledge makes this the obvious next step.
3. Positioning with Executives
Unusually, RPA is something of a silver bullet! limited up-front investment, limited training, and (sometimes) rapid results. As such, over time, it can lead to poor process selection and dilution of the potential in favour of ‘automating more stuff ’. Avoiding this requires prioritisation of opportunities across the greatest breadth of the organisation and devolution of authority to an ‘independent’ Centre of Excellence (CoE) which has clear objectives set from the top. Realising the value of RPA at scale requires ongoing adherence to the reasons why RPA was introduced in the first place–Cost efficiency, Customer Experience improvement, Risk reduction, and Revenue generation.
4. Positioning within the enterprise
Robots sit in Operations, at the junction of Customer, Data and IT. Process automation can be seen as an extension of labour arbitrage except that the labour is software! Successful RPA is a partnership between the Automation CoE and Business Process owners. Robots become part of the operations managers’ workforce, complete with on-boarding, capacity management, and process ownership responsibilities. Robots will fail, exceptions will occur. Business knowledge needs to be retained in the heads of increasingly knowledgeable operations professionals who better understand multiple end-to-end journeys.
5. Positioning with Staff
Automaton is a natural evolution of the traditional operations professional. It is good news for the operations teams as it provides the opportunity to incrementally move up the knowledge curve into more valuable and interesting roles. The ideal RPA expertise is a combination of operations knowledge, LEAN skills, agile change management, RPA, and technical architecture. These are the skills of the future operations professionals which combine existing and new capabilities. Furthermore, due to the transitional nature of the migration from manual to automated operations (as opposed to rapid transformation), the time is available for the operations teams to up skill as their organisation goes on this journey.
6. Next Steps - RPA and Machine Learning
RPA are mostly still ‘dumb robots’ which need to be told exactly what to do. The ‘magic sauce’ of RPA is not in the ability to execute a process; it is in the ability to interact with any system using the existing User Interface (UI). The natural progression is to make the robot smarter so that the laborious stepwise process definition can be simplified and to continue to avail the system interaction capabilities allowed using the existing UI. This is where RPA and Machine Learning should converge and where the overlap between RPA and AI will really emerge.
RPA is now an established capability which is increasingly being leveraged to automate operations across geographies and industries. Its continued growth and success will rely more on its ability to collaborate with the broader AI toolkit and on our ability to bring these tools together in a structured way. The future looks bright for RPA and for the broader AI landscape.