As Robotic Process Automation becomes more commonplace, we could well see a dramatic shift in the nature of human work in the coming years, as machines and algorithms get even smarter. But what are the true implications we need to be aware of?
Hollywood has been telling us for years that the robots will take over sooner or later. But those expecting some malevolent man-machine from an apocalyptic future – or an omniscient framework obscuring our true reality – to take their rightful seat in the next couple of years will be sorely disappointed.
But if technology continues to evolve at such an exponential rate, our daily reality could look quite different – especially in the workplace – perhaps sooner than we think. Yup, automation – or using machines to carry out repetitive tasks – has spilled out of manufacturing and into the office, call center – even the IT department. So the robots are already here, it seems. But where are they? Inside our computers of course! Or more specifically – in scalable cloud-based object storage.
What we are talking about here is Robotic Process Automation (RPA): an ambitious form of technology that is poised to change the way we operate at work in many different ways. But, to understand it, it is important that we map out exactly what we mean by ‘robots’, ‘automation’ and ‘AI’. We covered Business Process Automation (BPA) in a previous blog post, but to appreciate its more evolved form – RPA – its potential and how it works, we need to take a deeper dive into the terminologies, applications, and of course, the benefits and limitations.
The marriage of business and technology seems fickle at times, but the two have mutually beneficial needs that, when aligned, work very well indeed. Business Process Automation is clear evidence of that.
An offshoot of Business Process Management, BPA is essentially the fulfillment of a specific task using technology. It can be as simple as allowing an employee to request leave via a system and having their manager approve that request by email. Alternatively, it could be something as complex as a product development tool that allows engineers to collaborate in real time.
In each case, BPA allows users to easily alter information in a back-end system using a front-end interface. The keyword here is ‘users’. BPA empowers, streamlines, and enables task efficiency; but still requires human action. So, while its usefulness is evident; the ability to act independently is non-existent. And this is the major difference between BPA and RPA.
If BPA refers to ‘tools’ we can use to make work easier, then RPA defines the use of robots to take charge of those tools. However, robots in this sense do not (necessarily) mean there is a person-shaped piece of metal clicking a button time and again. These ‘bots’ may well be computer programs simply following a prescribed path set by their development master. Essentially RPA is the use of computing power to automate business processes – as dictated by structured inputs. It could be a business rule that issues an automatic response to an email; or a chatbot messenger service that responds to customer queries by serving them tailored information from a vast database of FAQs.
In technical terms, instead of altering a back-end process using a front-end interface, in RPA the information is both taken and input from the front-end – in the same way that a user would. RPA also allows information to be shared and stored across numerous different programs; with minimal supervision. Suffice to say it presents a marked improvement from what BPA has to offer.
However, RPA is not as big a jump from BPA as you might think. According to a recent Wall Street Journal article: “People with business process and industry expertise but no programming experience can start automating processes with RPA tools with only a few weeks of training.” Similarly, RPA does not require a huge system overhaul to begin implementation and can be used to carry out numerous tasks across a wide range of industries: from financial services to healthcare, as well as HR, marketing, and operations.
But is RPA essential? Should all businesses be sitting up and taking notice? And if not now, when? We shall take a look at the positives and negatives factors in more detail below, but overall, it is clear that there is a definite ongoing business need for technology to provide improved processes, more cost-effective labor, and faster operational speeds.
Is RPA the panacea many consider it to be? Well, it really depends on who you ask. If you are talking to those that are primarily concerned with time/cost/efficiency; the answer will be a loud resounding ‘yes’. Information and innovation leaders within businesses are RPA’s biggest advocates. Over 70% of the business leaders questioned by the Institute for Robotic Process Automation and Artificial Intelligence Survey said that they plan to invest more in the implementation and development of RPA in 2018.
But what does it actually offer? Automation is essentially the replacement of technology to complete tasks that would otherwise be carried out by humans. The average office worker has a lot of repetitive, routine tasks. So the more tech is put to work, the argument is that the time saving enables employees to focus on more meaningful work.
Switching out humans for robots also has other benefits. RPA means things get done faster and more accurately – and robots can also perform tasks at scale: any time of day or night. Add in the fact that automating tasks via RPA requires no coding and can be set in motion quickly, and the prospect of fast returns is clear enough.
Are brands actually making the most of what RPA has to offer? Indeed they are. Lots of very well-known ones: Walmart, Deloitte, Deutsche Bank, Ernst & Young, Telefonica, and American Express Global Business Travel are just some of the many enterprises adopting RPA.
Walmart already has 500 active bots; used to automate everything from answering employee questions to retrieving useful information from audit documents, while Deloitte LP redesigned its claims process deploying 85 bots to handle 1.5 million requests per year. This is the equivalent of over 200 employees – at approximately 30% of the cost.
Similarly, a four person team at Telefonica O2 in the UK, built 160 software bots to process between 400,000 and 500,000 monthly transactions; resulting in a three-year return on investment of over 650%.
Impressive though these figures are. We need to be clear about one thing: this kind of functionality does not mean the ‘robots’ are capable of doing anything other than what they are asked to do with the information they hold. RPA systems might be intelligent in a sense, but they are not capable of learning and improving. They simply perform – don’t they?
While there are plenty of tasks RPA bots cannot do as effectively as humans – such as exercising emotional intelligence, reasoning, judgment, and customer interaction - they are increasing in sophistication. Cognitive technologies like Machine Learning, speech recognition, and Natural Language Processing are making their way into automation; meaning that there is a strong possibility that bots will become mainstream customer service technologies.
In fact, in some ways they already are – and not just by taking care of the ‘boring’ or routine processes. As more customer-driven technologies and messaging services become more important service channels; the need for companies to be ‘always on’ increases too. Chatbots and Voice assistants may soon become indistinguishable from humans. And considering that online channels are some of the most important in the customer journey, being able to answer queries and offer support every step of the way, across numerous platforms will become an essential part of customer service.
According to a 2017 Forrester Research report, the RPA market is expected to grow from $250 million in 2016 to around $3 billion in 2021. However, that $3 billion is just a fraction of the $48.5 billion that companies have earmarked for overall AI spend. A recent study by Harvard Business Review of over 150 companies involved in AI-based projects found that over 70% of the projects were adding AI capabilities to their existing processes.
So if AI is the digital endgame for businesses, why bother with RPA at all? Because AI is still an unknown quantity for many – and RPA is a gateway to the use of Artificial Intelligence. Enterprises can supercharge their automation efforts by adding cognitive technologies – scaling their tech stack towards automating higher-order tasks.
By 2020 it is thought that automation and AI will top $1 billion; and reduce employee requirements in business shared-service centers by 65%. By then, some 40% of large enterprises will have adopted an RPA software tool as well.
Overall, while there is an overwhelmingly cost-effective, super-efficient technology solution to anything, the natural response is to jump at the chance of incorporating RPA into your business. However, its deployment should be carefully considered. It is important to keep in mind that enterprise technology, no matter how advanced it appears, is only useful when it solves a viable business problem.
In many cases, what automation does is use the information, insight, and expertise a business already has in increasingly intelligent ways. Without that kind of foundation, it will not work as intended. People will always underpin that foundation in any business. And until automation, in whatever form, can learn to develop and contribute to the way a company does business, how its people deliver on-brand customer experiences and provide service in the broadest sense; RPA will only ever play an auxiliary role in a business’ operations. For now.