13 August 2015
Robotic process automation: Are you ready for the second machine age?
The use of robots to perform human tasks has been the topic of numerous science fiction movies over the last century, though anyone who’s witnessed robotics on the modern car assembly line would hardly call it fiction.
Robotic Process Automation (RPA) has enabled the motoring industry to improve product reliability, enhance the quality of the manufacturing process and drive profound changes in its ability to produce a car to order.
By using software that mimics the actions of humans, tasks can be completed faster, more accurately, and at a fraction of the cost. The ever-increasing power of computing and evolution of smarter algorithms allows us to automate more and more of this activity.
Increasingly the line between RPA and automation in general is blurring. A more accurate, catchall title might be Intelligent Automation. Cognitive computing is expected to blur those lines even further as our desktop computers take over more of our routine tasks.
Process automation introduced at the RAC resulted in a
reduction in exception handling by back office support.
The introduction of automation by a utilities company resulted in an operational cost saving of
per annum over two years.
For example IPSoft’s Amelia is a ‘virtual service-desk employee’ that learns on the job, can reply to email, answer phone calls and hold conversations. IT infrastructure support jobs – help desks, data centres - are likely to be some of the first to feel this wind.
The business benefits of RPA are undoubted. Robotics helps improve profit and loss by reducing errors, avoiding costly off shoring and providing rapid results within weeks.
It also helps achieve operational excellence by improving quality and offering the ability to respond to changing business processes or handle increased volumes at marginal costs.
Furthermore, RPA helps a company leverage talent by giving employees time to innovate and focus on human centred activities such as customer service. Ultimately it gives business both the freedom and toolset to be creative and entrepreneurial. The things humans do best.
Here are just a few examples of where RPA can help:
- Pre-checking – run all case volumes through a process to perform checks, filter out the simple cases, let humans to do the complex work.
- New systems with missing functionality – if a system can’t be configured to do exactly what you want (or is too expensive) use automation to deliver the work. More cost effective than throwing people at it.
- Dirty interface – Gartner’s term for when lack of integration between systems means having to go through the user interface. In short, re-keying. Also true when input has to be provided to more than one system.
- Multiple sources of data input – again, if a process requires input from multiple sources then it is suitable for robotic automation.
- Manual checking, decisions and calculations – as long as these follow a rule set, then they can be automated. This is not about a technology that will take away our jobs, but a step change that frees us from mundane, repetitive tasks, increases our value and lets us innovate in new areas.
Chasing artificial intelligence
The collection of facts an ordinary person knows. The AI field of knowledge representation aims to build a database of all the general knowledge most people possess. It’s colossal, one of AI’s biggest challenges.
SI systems comprise a population of simple agents that interact with each other and the environment. Over time intelligent behaviours emerge that none of the individual agents understand. Think ant colonies and bird flocking.
The Holy Grail. The term used for a machine with general intelligence, not only capable of all the learning and understanding capacities of a human but able to exceed many or all of them.
What our experts say
Introducing robotic process automation (RPA) can facilitate a significant improvement in output quality in situations where IT input and reporting processes currently require considerable manual intervention. However, RPA is not necessarily always a direct replacement for other outsourcing solutions such as off-shoring. Areas suited to the application of RPA need to be identified rigorously, as not all processes are ideal candidates, but in many cases we have seen over 80% of transactions automated and cost savings of 70% made against manual alternatives. Good examples include the repetitive processing of data from multiple systems such as that required in certain Council Tax collection processes and the provision of illustrations for long term insurance policy holders. What has experience shown us? Well, careful planning and design are crucial. No organization has the same legacy system and process mix, and although the tools are generic, bespoke design is needed to reflect local requirements.
Solutions director, Capita