Robotic Process Automation – Fact, Fiction and Opportunities

Introduction of computers in banks and mechanical robots on production floors seem to have happened a long time ago. These concepts of automation revolutionized industries forever and set new benchmarks in terms of efficiency and quality.

Robotics Process Automation (RPA) is a software industry’s counterpart to production line robots. Present decade has seen evolution of cutting edge technologies – Artificial Intelligence (AI), Big Data, Internet of Things (IoT) and Machine Learning (ML) – which can be harnessed to deliver unmatched quality in a cost efficient manner, saving considerable human time.

Information Technology automation has been a common boardroom topic for years now. It has many success stories to its credit, where enterprises have shown the foresight to invest in automation to gain a competitive edge over competitors.

Automation, in its traditional sense has been heavily dependent on structured data processed through heavy scripting languages to deliver the results. It gets the job done, but it has faces performance issues.

RPA is the capability of a software or service to perform rule based IT tasks the same way a human would. This capability can be extended to automate complex rule based assignments. Imagine the work floor of a credit analysis firm – an analyst would consume data from a myriad of sources to determine an applicant’s credit history. Once done, he would take a decision pertaining to loan sanction, processing fee waiver, determining a suitable interest rate, etc.

During the process, he would log into multiple application, scrape out information from websites, run spreadsheet based models and execute other tasks to arrive at a decision. RPA capability can be used to automate this task and save his time for more valuable assignments that require human judgement. These qualities make open up immense possibilities in tasks related to banking, accounting, compliance, audit, customer support and record keeping.

Here is a ready reference list of RPA capabilities that can help you decide whether or not an RPA managed process is worth the investment:

• Opening e-mails and attachments
• Logging into Web Applications
• Move files and folders
• Scrape website data
• Connecting to system Application Programming Interfaces (APIs)
• Follow If-then decision rules
• Extract data and build dashboards
• Parse documents and extract structured data
• Social media statistics aggregation
• Aggregate data hosted in multiple systems and sources
• Calculations
• Copy-past data
• Form filling
• Read and Modify databases
• Manipulate applications

Alex Edlich and Vik Sohoni, senior partners at McKinsey & Company, find, “several robotics programs have been put on hold, or CIOs have flatly refused to install new bots.” Before we dive deeper into RPA and determine how it can be used beneficially to save time and costs, we need to understand what RPA isn’t. At present it is difficult to imagine an enterprise which hasn’t already adopted RPA or isn’t seriously thinking about it. As more and more organization continuously join the RPA bandwagon and expect enhanced capabilities and accuracy, we need to see through the technological clutter and see its limitations up front.

In the previous section, we mentioned RPA is essentially a software program, popularly known as bot. The term “bot” sparks an image of a walking and talking robot that can move around our facility and perform a multitude of tasks tirelessly. That is exactly what RPA cannot do on its own. RPA is not a physical machine that can move around or process actual papers. Nor it is your AI powered assistant right out of a science fiction movie that can be voice programmed to listen to your queries and respond instantly.

These terms are often used interchangeably (and inaccurately!) and it would be worth the effort to try and distinguish the two. Primary differentiator is that RPA can be viewed as an end-user oriented solution. It operates at the foreground and goes about its tasks just like a human would – logging into different applications, picking up data and processing it based on pre-defined rules.

To put things in perspective, RPA can replace an office worker who spends 3-4 hours a day performing routine tasks such as data entry, batching, transaction processing, account settlement, etc. The employee can invoke the bot whenever tasks build up in the pipeline. Processing is much faster, more accurate and it leaves the worker with a lot of time on his hands to focus on activities that add more value.

IT Process Automation (ITPA) has been around for a while now. It essentially operates in the background, much different from mimicking a user’s workflow. Consequently, it is more complex and takes a long time to be developed. It can be used effectively in Change Management and Incident Management processes, where an incoming ticket serves as a trigger.

As RPA is a programmable solution, it can be configured to automate a variety of tasks that consume days and months for operations employees. Some of the examples are calculations, record maintenance, transaction processing, queries, data entry, etc. RPA software, commonly known as bots, are becoming more and more sophisticated using technologies such as screen scraping, workflow automation, Artificial Intelligence and Machine Learning. This enables it to deliver high quality results in an efficient manner.

Unlike traditional automation solutions, RPA based solutions have a lower time-to-market, thanks to its fast development and implementation cycle times. In addition, you do not require extensive coding skills to learn to automate processes using RPA. Your general workflow designed in Microsoft Visio can be used as a baseline to get started on RPA tools and end users can be trained in a short time to come up with detailed descriptions of their processes, leading to automation.

Some of the most common benefits of RPA that can be leveraged across industries are:

• Enhanced customer support
• Improved audit compliance for business processes
• Rapid turnaround
• Cost savings
• Improved accuracy and reporting
• Zero pipeline

It is very easy to have an information overload on RPA on the internet and lose track of the information you are seeking. At the end of the day, business processes identified for RPA have to justify the investment through improved input costs, efficiency, reduced cycle times and near flawless accuracy. Here are few characteristics to look out for while identifying processes that are ideal for RPA investment:

• High operational cost process – staffing of 5 or more
• Repetitive, rule-based processes that require manual access of one or more systems
• Processes interacting with structured data sets
• Accessing applications hosted on windows, web platforms and mainframe
• Processes involving higher transaction volumes
• Processes involving long cycle times
• Highly regulated processes requiring zero-error execution
• Manual data entry prone to errors

When you’re running a business, you need to crunch the numbers to justify technology adoption. As we have indicated in previous sections, you need to run your targeted processes through a checklist to see how many boxes it ticks – in order to determine its suitability for RPA. Here, we present general guidelines which would indicate the sophistication level of RPA processes and its related cost savings.

The following table, published by Ernst and Young in their RPA Whitepaper in 2015 maps RPA sophistication levels to the kind of tasks handled and associated cost saving estimates:

Level	Types of RPA technology	Description	Estimated cost savings
Class I	Basic Process Automation	Macros, screen scraping and business workflow technologies in the presentation layer; not integrated	10-20%
Class II	Enhanced and Intelligent Process Automation	Technologies using natural language processing; able to understand unstructured data and apply it to process automation	35-50%
Class III	Cognitive Platforms	Cognitive computing systems that essentially attempt to solve problems in the same way as humans, by learning from experience and acting on that learning	> 60%

All that information on RPA would come to nothing if it did not have proven cases of huge savings in cost and manual time, would it? In this section we pick up cases from some of the top organizations around the globe – from various industries – and see how RPA helped them overhaul manual processes leading to enhanced efficiency and accuracy.

Manchester, United Kingdom headquartered The Co-operative Bank places utmost focus on providing excellent customer service. It conducts periodic reviews of its processes and identified 10 processes to be automated within a 12-month time frame. RPA pioneer Blue Prism helped them achieve this target with great efficiencies, freeing up employee time for more customer facing roles. Entire case study can be found on Blue Prism website. Some of the key takeaways from the case are:

• 10 processes identified and automated within 12 months
• Audit conducted in 1 minute with automation versus 6-7 hours manually
• 200 accounts per hour closed automatically versus 12 manually
• CHAPS process reduced to 20 seconds automatically versus 10 minutes manually
• Rapid ROl – individual projects paid for in less than 3 months
• Significant FTE saving (full time equivalent salaries) in Year 1
• Improved levels of customer service
• Significant improvements in speed, consistency and accuracy of processes
• Processes built by the business users – reduced IT cost to support project

Steve Chilton, University Hospitals Birmingham’s Deputy Director of IT services, took up the challenge to build self-registration systems to handle large volumes of patient inflows. The challenge was to synchronize multiple processes and integrate them on a common technological platform in order to reduce manual intervention and improve efficiencies in the long drawn process of patient check-ins. Partnering with Blue Prism, the hospital introduced first of a kind self-administration kiosk in NHS Trust Hospital, leveraging National Program of IT (NPfIT) patient Administration Service. The results were extraordinary in terms of cost, efficiency, accuracy and end user satisfaction.

Some of the key takeaways are:

• 50% increase in staff efficiency
• High adoption rate of 51% among patients
• Zero que, easy to understand user interface resulted in high patient satisfaction ratings
• CHAPS process reduced to 20 seconds automatically versus 10 minutes manually
• Self-registration saw twice as fast patient flow compared to manual process
• Data quality improvement
• Less than 3 months’ implementation cycle
• Enhanced integration with Patient Administration Services
• Better adherence to security and compliance policies

Once you identify your processes that are candidates for RPA, you need to choose from a number of service providers playing in the industry. Choosing the right software is as important as automating the process itself, as you do not want to end up in a situation where you have to follow up with your RPA solutions provider for every little modification and customization. Here are few key features to look out for to ensure your RPA adoption is seamless, always-on and interruption free:

• Hosting – Can the solution be deployed on Cloud, customer terminal as well as Virtual Machines? These features help adoption and scalability for enterprises operation in a thin client environment.
• Compatibility across Virtual and Enterprise Applications – A software’s proven track record of integration with enterprise applications such as SAP, Siebel, PeopleSoft, FIS, BMC Remedy and virtualization technologies such as Citrix and Xen will go a long way in unlocking full potential of RPA.
• Release Environments capability – The solution should enable a role-based actions to be performed in different environments, i.e. testing, staging and production. Depending on a tester’s observations, a workflow released by a developer in staging environment should be sent to production servers (successful test) or rolled back (failed test).
• Rapid development support – A large number of modular, re-usable features would go a long way in facilitating shorter development cycles when a working solution is to be customized for different environments.
• Scalability – An ideal RPA solution should be able to up-scale and down-scale hundreds of robots simultaneously, depending on workload. It should have a web based management console, from where its operations can be controlled and monitored.

The market for RPA is immense and given the rapid strides being taken in the fields of Artificial Intelligence, Internet of Things (IoT) and Machine Learning, application of RPA is limited only by one’s imagination. Rather than looking at it as a technology that has set out to snatch away human jobs, look at it as a technology which assists humans in performing routine tasks, freeing up time to take up more challenging roles that add value. If you are looking out for an RPA consultant in Ireland, contact us and we will help you unlock the full potential of RPA. Happy automating!