When it comes to present-day automation and repetitive work, we have traditionally thought of actual physical machines, or an entire line of robotic arms and tools that process items from one end towards another.
Today, however, we are now getting used to yet another, more virtual type of automation. This is none other than the concept of robotic process automation (RPA): the automation of processes based directly on software and artificial intelligence (AI).
But how exactly can we define RPA? We take a look into its brief development history and a few real-world examples, to see how it works, as well as witness its impact on organizational productivity as a whole.
Brief History of Robotic Process Automation
Vestiges of RPA started way back when the business/consumer computer market started to take off during the 1970s. This was mostly just limited to the computerization of standard end-business processes. However, during the 1990s, tech giants such as IBM started replacing basic, paper-based processes with electronic tasks using the first generation of workflow automation software.
As the dawn of the 21st century arrived, the first true robotic process automation systems were finally developed. Blue Prism, UiPath, and Automation Anywhere introduced the idea of step-by-step automation, where each manual step of any software process can be preserved, which would then be tweaked to become fully automated tasks.
With the further advancement of machine learning technologies today, RPA is taking the next steps towards yet another development revolution. User input, a standard requirement to constantly set up the software, can now be potentially eliminated in favor of “unassisted” systems, or the independent development of processes without any user intervention.
General Applications of Robotic Process Automation
For our exclusive definition within this article, we shall divide the general uses of RPA into two familiar office categories: front-office, and back-office applications:
Front Office Applications
- Customer service basic queries – use chatbots to cater to FAQs
- User access guidance – step-by-step navigation through heaps of information
- Customer info organization and storage – automated database entry and notification
- Sales order and invoice processing – unassisted sending of sales data to related personnel
Back Office Applications
- Employee data management – action/activity recommendations via points of interest
- Database organization – automated chiving, sending, updating, segregation, all-in-one
- Pricing and competition intelligence – organized delivery of info on other companies
- Data aggregation and management – collecting related data, pre-set pattern recognition
RPA use cases can go well-beyond commercial applications. Any regular or repetitive process that deals with any information in a non-creative way is most likely a good candidate for an RPA system. However, for the most part, the technologies associated with RPA provide the most benefits within the business sector. As such, a significant portion of our use cases will always involve standard or commonly known business practices.
Engineered Productivity: Blue Prism, UiPath, and Automation Anywhere
To elaborate on what we have explained about RPA so far, we need to briefly introduce the three most popular RPA tools function for a typical user. From a general standpoint, these software options employ what they call an automation workflow. Basically, they introduce a suite of features that help automatically do what otherwise have been considered mundane, repetitive workstation tasks.
However, do take note that despite having the same end-objective, the methodology and design principle of each of these modern-era RPA solutions can differ considerably. The difference can be based on the level of automation, processes involved, and the integration of the user’s knowledge about programming and software development in general:
- BluePrism – externally feels like the more corporate, purely operation-oriented solution of the three classic RPAs with its back office-only solutions. It has options to allow the automation of either desktop actions, web or Citrix-based processes, and offers a mixed variety of visual editing tools and script-based (programming) setups. Unfortunately, there is no way to test Blue Prism personally unless you purchase their software.
- Automation Anywhere – currently listed as the least popular of the three classic RPA solutions. It is more accessible than Blue Prism, having the option to test the software first for free in its Community version before letting the user go for more enterprise-level options. There is a significant requirement for programming knowledge with this RPA, as well as being exclusively Microsoft-based. Both of which probably contributes to its lesser popularity.
- UiPath – the most popular out of all the three classic RPA solutions. Is it the best? Structure and feature-wise, no. Automation Anywhere can usually provide just as much functionality. But the convenient combination of an accessible interface, multi-platform options, zero programming requirement, and flexible (highly modular) base-level workflow automation makes it the choice of many starting RPA users worldwide.
Real World Use Case 01: Royal Bank of Canada’s Service Expansion Chatbots
Royal Bank of Canada is one of the Big Five banks of its respective country, and as such, is an integral part of the nation’s economic backbone. Around three years ago, the finance giant started heavily investing in chatbots, or human conversation simulation programs. This is to help significantly expand their basic customer service systems, with the secondary objective, of course, to reduce costs via automation.
As presented in a demonstration at TechToronto last 2017, the chatbot can automatically deal with basic visitor guidance instructions for their official websites. Instead of manually navigating through pages and pages of potential services, users can interact with a dynamic automated response system, which would then promptly try to answer simple questions.
In fact, depending on how long the interaction has taken place (how many questions were answered), the programmed chatbot can provide increasingly complex responses. Ask enough questions about finding a residence long enough, and it will eventually give more pointers on how loan applications are processed. Though, of course, the end objective is still to redirect users to other website sections where the desired information may be located.
“In theory, that means the human user can be freed up to more judgment-based creative tasks. [Though] in reality, it often means that human users are reduced and turns into cost savings for the bank.”
-Joel So, financial partner (PricewaterhouseCoopers Canada)
By using an RPA to automate responses, the Royal Bank of Canada maintains its competitive economic edge. Even better, at that specific time, at least two more of Canada’s Big Five banks were heavingly investing into RPA. For example, Toronto-Dominion Bank was also developing social media-based chatbots, in order to cater to a portion of their customers who may be visiting their respective social media pages. This consequently provides more room for development and growth via competition, as well as the opportunity to integrate RPA even deeper into the company’s routine systems.
Real World RPA Use Case 02: UiPath’s Very Own Brew of Personnel Optimization
Significantly enhancing HR management is also one of the repetitive tasks that RPA can automate. Most of the benefits revolve around the concept of indexing employee databases and connecting event-triggered tasks for various purposes. A few notable ones include managing personal schedules, analyzing payrolls, setting up automated notifications when certain conditions are achieved, among many other things.
UiPath was one of the earliest developers of RPA tools as we know it in the modern era. As such, the company definitely knows how to optimize its technologies to enhance human resource management. In a brief official blog post by the company, analyst Robert Stewart described how an RPA could be essentially used to remove the very HR element itself, by replacing those slots with virtual office robots. The article then further explains how it is possible to reallocate employees, where their critical thinking and analytical skills matter the most. In principle, this essentially increases potential productivity in two separate directions.
“It just makes the job better. I can focus on other projects or processes that are more valuable. I can focus more on interacting with people or thinking about Corporate Social Responsibility (CSR) projects.”
-Mari Ambrose, People Operations Leader (UiPath)
Needless to say, other less frequent (but still quite important) mundane tasks, such as managing employee absence time (sick leaves, vacation requests, etc.), can also be automated. Got your employees stuck all weekend dealing with expense calculations and reimbursement processes? Get an RPA to automate the task and set a manager that will monitor its activity.
With a robust and frequently tested deployment of the right RPA, data and process errors can be kept at a minimum, at least where the human element can simply intervene to provide the necessary corrections.
The Challenge to Keep RPA Beneficial
However, as revolutionary as RPA is in any modern, corporate setting, there is one thing that still hinders its universal adoption: maintenance. Designing the automated software elements (bots) is just but one of the steps in successfully using an RPA for your organization or business. There are still several phases that needed to be addressed that is related to its maintenance, such as:
- Systems Testing – this phase is primarily implemented to check if the RPA build works in the first place, although this is also important to determine bugs and errors.
- Deployment – the official implementation of the RPA. This is also where integration to current systems take place, a step that might be a lot more complicated depending on the technologies already available by default on your system.
- Support – After deployment, additional updates may be necessary to patch some issues or append unforeseen necessities during the design phase. Basically, this phase attempts to ensure that your RPA still provides the benefits it was built to achieve.
As a matter of fact, Nasdaq even acknowledges this inherent issue with RPA, with a survey report last year that concludes the difficulty of deployment and maintaining RPA. The report stated that many companies had several misconceptions about RPA’s potential, and how it is perceived when it is finally implemented.
The general recommendation was that RPA should be treated as a framework, or a bridge, not an absolute “endgame” determining factor. And like all foundations, support and maintenance are required to keep its intended function as it increases in scale.
Robotic process automation is indeed a significant game-changer in the world of business and commerce today. While its philosophical arguments about giving more focus to human creativity are true to an extent, the fact of the matter is that an RPA system would always be more on the side of optimization. This isn’t just about lowering costs by eliminating human resources, but this also about creating a completely new way of organizing business information.
There is a caveat to all this, of course. Good knowledge of RPA’s technologies is usually required before you can even attempt to create a decently competent RPA. But, the process of learning something new is a sacrifice every one of us should be willing to take, no matter how relatively outdated our age might seem compared to the ever-evolving world around us.