Saturday Nov 23, 2024
Thursday, 14 July 2016 00:00 - - {{hitsCtrl.values.hits}}
Panel discussion Moderated by ICTA Program Head Fayaz Hudah. Others from left: Managing Director Hayleys Industrial Solutions Dr. Arul Sivagananathan, NTB Chief Operating Officer Tilak Piyadigama, Chief Operating Officer VirtusaPolaris Keith Modder and Global Head of Engineering VirtusaPolaris Chandika Mendis
By Madushka Balasuriya
The Industrial Revolution and the dawn of the internet, two of the most ground-breaking advances ever made in human history. Is Robotic Process Automation (RPA) the next? That was the question posed at a CIO Forum organised by VirtusaPolaris in association with the Computer Society of Sri Lanka (CSSL) and the Federation of Information Technology Industry Sri Lanka (FITIS), provocatively titled ‘Software Robots: Friend or Foe? The choice is yours’.
“Sometimes in our history we get to a point where a very strong advancement in technology intersects with a dire business need. When these two meet, the outcome is a paradigm shift,” stated Global Head of Engineering VirtusaPolaris Chandika Mendis, at the top of his keynote address.
This type of paradigm shift, he elaborated, involves steep increases in productivity, quality, user experience, and “at times even the pace of innovation itself”.
“If you look at history, every time this kind of paradigm shift has occurred there has been a very big shift in the job composition. It’s not that jobs have gone away – jobs have increased – but the nature of the jobs have changed drastically.”
“We are now at one of those points in history,” he continued, pausing for effect. “Software robots are the new revolution.”
What is Robotic Process Automation?
At its core, RPA is “robotic” software that organisations configure to capture and interpret the actions of existing applications employed in various business processes. Its aim is to automate a variety of business and computing processes typically handled by humans, and due to ever-increasing advances in software and artificial intelligence, this automation has its sights set on higher-value tasks than ever before.
Once RPA software has been trained to understand specific processes, it can then automatically process transactions, manipulate data, trigger responses, and communicate with other systems as necessary. The technology is designed to reduce or eliminate the need for people to perform high-volume IT support, workflow, remote infrastructure, and back-office processes, such as those found in finance, accounting, supply chain management, customer service, and human resources.
“Most of the actions that human operators are tasked with have got streamlined. These actions are now very repetitive and structured, which has increased efficiency in some ways. But this means that these tasks are very easily accomplished by a software robot,” pointed out Mendis.
“Human beings are not wired to do this repetitive work. They’re wired to be creative.”
Why RPA?
In his address Mendis also drew attention to a startling figure: While globally the manufacturing sector has seen productivity increase by 58% since 2009, the services industry – such as Banking, Financial Services and Insurance – has seen an 8% drop. This he attributed to the increased use of robotics – not to be confused with software robotics – in the field of manufacturing, and the comparative lack of use of automation in the services sector.
When it comes to service companies they usually provide a single interface for their customers, where a human operator bridges the gap between the customer and the variety of different IT systems which handle their requests (this is usually done through a single computer). This system has become such an integral part of the services industry that it has given rise to an entirely new industry – Business Process Outsourcing (BPO) – as more firms find it increasingly cost efficient to outsource customer service-related tasks.
As a result “competition in the services industry has been intense and newcomers have introduced newer, leaner products, and are taking over market share,” explained Mendis. “So the big boys have had to adapt. What that means is more complexity in their service product mix and more complexity in the product overall.”
However with the nature of developing new IT systems a complex, expensive and time-consuming process, most systems have been unable to accommodate this shift. Instead human operators have had to undertake ever more complex tasks to best utilise their given systems. This is less than ideal, as human operators are error prone; require extensive training to operate these systems; and the work they do is not easily scalable.
“Speed is the operative word for businesses. IT systems have just been unable to deliver. This is the reality. And this is true throughout the industry.
“The moment you automate the process, you completely streamline your scale. Bots operate 24/7, and you can always increase the number of bots if your peak workload increases. We’re not talking about programming here. We’re talking about mimicking human actions. It will work on the same interfaces human operators work on.”
In addition to completing a human operator’s tasks, the robotic software can also be superimposed on to IT application program interfaces (API), which are a set of routines, protocols, and tools for building software applications. Mendis promised that automation in this form will be faster and simpler than through programming.
“All you need is someone who can encode the necessary operations in the programming software. Whereas IT systems have month-long installation cycles, a typical RPA installation is weeks; some we have implemented in days.”
“We are talking about a huge amount of power to automate any actions done by human operators; repetitive, time-consuming, expensive actions.”
Real world implementation
While relatively new to Sri Lanka, RPA software has been making inroads in various sectors globally, with several examples readily available of how its implementation has disrupted the industries it has been introduced to.
A bank in the United Kingdom managed to reduce the cost of processing high risk accounts by nearly 80% after RPA software was introduced to assist 11 full-time employees (FTE). This in turn allowed those employees to redeployed in more proactive customer related activities.
Similarly a major global bank automated their services for fraudulent account closure, loan application opening and right of ‘setting-off’. This resulted in the work of 120 FTEs reducing and reduction in bad debt provisions of £175m per annum.
Meanwhile in the field of retail manufacturing, automation allowed a firm to reduce the time spent on extracting data from various Excel and PDF inputs from five hours per order to just two minutes per order. This resulted in up to 95% worth of savings when it came to pre-order processing, as well as complete elimination of the high cost of errors.
Finally a large US firm introduced a Service Desk robot that was able to answer customer calls, which ended up with the bot answering 62,000 calls in total. And two thirds of those calls were successfully completed, with the problems resolved, all without the intervention of a human operator.
Addressing concerns
However, despite these seemingly overwhelming positive examples of employing RPA software, human beings are by nature resistant to change. In a question and answer session towards the end of the forum the audience had the chance to put their reservations to a panel of experts, some of whom had reservations of their own.
Work experience
One member of the panel was Managing Director Hayleys Industrial Solutions Dr. Arul Sivagananathan, who by his own admission is not an early adopter of new technology. His scepticism was centred around the concern that employing bots would diminish the problem solving capabilities of new recruits, and deprive them of an essential learning curve.
“The most important thing when we talk about finance and accounting, if you get a junior who is 18 of 19 years of age coming in you can’t put them straight away into middle management,” explained Dr. Sivagananathan.
“Rather than just using a system which says that this is the outcome, they need to understand what that outcome means. We want someone to learn the process as they go up through the ranks, where they’ll be able to add value to improve the process. That’s what we’re looking for.
“I’m not saying a shift into software robotics won’t bring the cost down, but at what cost? We need to have processes which educate people so that they understand what the data means.”
Despite his concerns though, Dr. Sivagananathan went on to explain how he had seen first-hand the benefits of automation software, where Coca Cola over the space of 15 years went from having 30 employees performing and accounts payable job to one person handling it over the course of half a work day on their mobile phone.
“Yes, I have seen this evolution, I encourage this evolution, but it needs to be done in a way that the younger generation that comes in does not lose out on the learning curve we have gone through,” he concluded.
Loss of jobs
Another concern was the prospect of RPA software eating into traditional IT infrastructure support jobs. Chief Operating Officer VirtusaPolaris Keith Modder, looked to abate those concerns using historical examples of industry wide disruption. “About 20 years ago about 80% of the revenue from the BPO industry was associated with data capture. 20 years since they don’t do any data capturing because optical character recognition and scanners have taken over. And that industry has grown twentyfold.
“And I do believe that as organisations reduce their cost/income ratios, those workers and team members can be adopted to provide much higher value added services. The need for data scientists and data analysts is going to be a lot greater.”
This was echoed by Mendis.
“RPA does not eliminate the need for process specialists. It might replace a lot of people who are doing mundane data entry/analysis tasks, but it won’t replace the person who understands those tasks,” he explained.
“Those are the specialists who have functional knowledge and banking experience; they become much more impactful because of RPA, they then are able to start worrying about how to improve customer experience, cost-efficiency and so on.”
Hyper-competition
Another point of discussion was the rise of hyper-competition. Due to the ease by which RPA software can be implemented (three to six weeks), there was concern that any perceived edge a company might gain over a competitor would be short-lived. Mendis was adamant that this would not be the case.
“RPA does not eliminate the need for IT systems that we have today, and one of the challenges that we have with those systems is keeping up with the pace of change. What RPA does is really bridge the gap in those systems that arise from the integration of human operators. So companies are still going to be having an advantage based on the core system they’re having.
“I think the key thing is going to be how a company adapts their RPA system to really be client centric, how they innovate around that. Just because you’re RPA capable doesn’t mean your company is geared towards innovation. That requires a different mindset and a different level of leadership.
“There’s a lot of scope for improvements on the IT systems that RPA software will be running on. What RPA is offering right now is almost a band-aid solution. There’s still a lot of scope for disrupting the way we develop, maintain and evolve software.”
What next?
RPA is most likely to replace data entry and formatting tasks, which are rules-based. Earlier advances in automation eliminated blue collar jobs, and RPA will shift the focus of automation up the value chain. At the same time, demand for IT experts is growing for even higher-value jobs. This is our present reality, but what does the not-so-distant future hold for software robots?
One factor that will play a major role in answering this question is the increasing amount of research and funding being put into artificial intelligence (AI). AI has been making waves over the past few years with major tech players such as Facebook, Google and Microsoft funding in-depth research into the field. However the software is far from perfect. Microsoft recently had to take down an AI chat robot that it introduced to Twitter after it transformed into a Hitler-loving internet troll. This was down to the fact that the robot’s responses were learned by the conversations it had with real humans online – or in this case the data it was fed.
Going forward, as AI technology advances protocols and standards around bots will need to be formalised but for now, Mendis explains, “this technology is still only couple of years old, and the protocol itself is still vendor specific”.
When it comes to RPA software in particular, Mendis noted: “There is an artificial intelligence aspect that can be added on, such as speech recognition and character recognition or actually learning the actions of the human operator.
“But even without that there is a huge usage in terms of RPA. I think in the future this kind of system will have a variety of uses depending on the data that you feed it. Human work flow and business process management (BPM) has its place, yes. But RPA combined with BPM provides the biggest benefit.”
RPA is often called the automation of automation, and with good reason. According to the Institute for Robotic Process Automation, RPA can create a 25-50% cost savings and its process automation enables 24/7/365 execution at a fraction of the cost of human equivalents.
Typically referred to as ‘digital labour’, a software robot can cost as little as one-third the price of an offshore full-time employee (FTE) and as little as one-fifth the price of an onshore FTE. On top of that software robots can be deployed on existing IT infrastructure which means it adheres to existing regulations as setup by the IT department.
To quote Mendis: “Imagine the disruption that’s going to make.”
Pix by Lasantha Kumara