Robotics Process Automation

“The best way to predict the future is to invent it.”

What is Robotics Process Automation or what’s RPA? This is the question most of the people ask me when I say I am a Robotics Engineer. It is a Software robot design to replicates human interaction with different user interfaces in order to complete a process or a task. It can be a transaction submitting to an ERP system (SAP Business Object, Oracle), it can be data collecting to create different types of reports.

When I give this answer, they say “it is easier for you to develop these kind of things since you are a programmer.” The difference of RPA is that you do not have to be a programmer or a developer or even a software engineer to design and develop RPA bots. There are RPA platforms in the current market and it has the feature to create the logic of the process just by drag and drop stages to create the process flow. Automation Anywhere (AA) and Blue Prism is the most popular platforms.

Where we can use RPA? The general characteristics of a process where robotics process automation can benefit more with the processes which is repetitive and rule based, transactions on applications which can be a windows or web based platforms, data input is prone to human, not fragmented processes and more of back office processes that outsource to offshore locations which can be Finance & Accounting, Reporting, Banking Insurance, Telecommunications, Healthcare, Retail, Government, HR and IT.

Robotics Process Automation market is growing rapidly transforming businesses and it forecast the growth to be 20% in 2024 increasing the market share exceeding USD 5 billion as for the new research report by Global Market Insights, Inc. RPA platforms already been partners of Google, KPMG, Accenture, Appian, IBM, EY, Microsoft, Capgemini, Deloitte and it uses more than 700 global enterprises Dell, CocaCola, O2, LinkedIn, Whirlpool, ANZ, AT&T, BNY Mellon.

What my experience in RPA? For the past 1 and half years I have been working on Blue Prism continuously driving RPA projects in Agile Methodology and we are the first to build a RPA bot from entire WNS Global Services. I would say it was not an easy journey since I had to self-learn RPA by trying new things, learning from mistakes and research on internet for knowledge. Journal Automation RPA bot is one of the greatest achievement we developed covering 3000+ journal preparation automatically from opportunity identification, Data gathering and analyzing, Solution designing, UAT testing and Deploying and Maintaining. This is an unattended bot completed the journals end to end including the journal input extraction, input validation, journal creation by different calculations, journal review and upload and post to ledger. The processes identified for this solution is very fragmented and because of that the approach we took, first gathered all Journal details with different parameters source input, similarity of the journals, complexity, applications interact, source input type then conducted walkthrough and If possible eliminate the process (e.g. by doing hub fixes) then Reengineered/ Standardized the process and Solution developed for automation.

The benefit of using RPA is that the solution runs 24/7 automatically trigger in production region, exception trigger taking different steps (e.g. sends mail on exception reason), 100% reliable output, SPLUNK report generates on all stages creating MI’s, Non-disruptive, Efficient, Error free and Cheaper.

The future of Robotics Process Automation is the Cognitive stage where it uses machine learning and artificial intelligence. What is Artificial Intelligence or Machine Learning? A computer programmed to learn from experience to perform decision making, and natural language processing and pattern recognition. Why we need AI or Machine Learning? Human brain can perform unlimited task and it learns from experiences and knowledge it gathered. Sometimes there are specific task which we do not know until we try it. So as a programmer how can we write a program or even create a RPA bot if we do not know how to do it at first. And even if we know it will be a customized solution only for that specific task. Instead of writing high complex algorithm for each specific task, we collect lots of examples that specify the correct output for a given input. Then machine learning algorithm takes these examples and produces a program that does the job. How do I know this? Because I am coming from an Artificial Intelligence background.

When technology continuous to develop we need to adapt to the changing world in order to survive in the market. Since RPA can work 24/7 with all these new features better, faster and cheaper, who wouldn’t invest on it?