This article is published on LinkedIn on Feb 8th, 2021 by Mr. Ravi Ramamurthy – CEO, Epiance
In the late 19th century a few prominent physicists in a fit of hubris proclaimed that we had reached the pinnacle in Physics and there was nothing more earth-shattering left to uncover. They lamented that humankind would leave the job to the jejune to ferret out the discoveries and convert them into mundane applications.
A few decades later, a little known physicist, Albert Einstein upended the community with Quantum Science and Theory of Relativity’s twin discoveries. When a society is at the cusp of grandioseness and boldly declares the arrival of perfection, surprises displace them from their positions of zenith. One specific technology where I see such hubris creeping in is RPA (Robotic Process Automation). Just a few years back, there was intense excitement about the opportunities presented by the technology and today I hear that many of them have reached advanced levels of maturity. When one digs deeper to understand the prevalence of automation, I hear a retort with pride- ” We have deployed 100 Bots” and that in an organization which has 100,000 employees.
The truth is that the RPA revolution has just begun. It has just left the closet of pilots and proof of concepts. It is also true that the band of applicability of RPA has been too narrow. Organizations have chosen the most straightforward problems, those scenarios where the process can be automated end to end by an unattended system. Opportunities for automation exist in numerous use cases. BOts have replaced just a minuscule of routine work. A bulk of everyday activities are more complex, involve partial human efforts or require customization. Despite the advent of AI and Intelligent Automation, only a few organizations truly understand the import of these terms. Converting these concepts into practical, real-life applications in a scalable and sustainable manner has been a challenge. As usual, the hype has run faster and has overtaken the reality on the ground.
To summarise the challenges concerning implementation have not gone away, the technology has not matured, and automation opportunities have not been exhausted. One has just touched the tip of the iceberg.
Three changes are required to usher in the next big revolution in RPA.
- Reduce time and efforts to create BOTs: The creation of BOTs is still very tedious and time-consuming. The promise of drag and drop process for creating bots exists only in the realm of wild imagination. Availability of experienced technologists has become essential to ensure that the right process is selected, the solution is architected correctly, and the BOT’s creation follows best practices. In the absence of such bulwarks, Bots’ creation has a striking similarity to a software development process. Inexperienced and mediocre resources create lousy code, and inexperienced BOT developers create substandard BOTs. RPA requires individuals with a deeper knowledge of technology than was initially anticipated. Business analysts cannot create Bots. We know that now. Once we enter the confines of a waterfall framework of development, the time from conception to rollout stretches to weeks and months. In many situations, the underlying regulations, policies change, and the base application itself changes. The BOTs then become obsolete before they see the light of the day. The longer the time more are the uncertainties and lesser is the return on investment. BOT creation has to become more automated for it to become viable. Development environments (.NET, Angular, or React) have to embed RPA constructs to enable one to embed RPA functionality. Organizations should automate the actual process of BOT creation itself, cutting out many of the time-consuming activities( such as Process understanding, Business Requirement document creation, writing the script etc.).
- Enhance the scope of RPA: RPA should coexist with human beings. Companies should create a seamless environment for BOTs to hand over a task to human beings and vice versa. Framework and technologies have to evolve that will facilitate such functioning. Attended BOTs, BOTs and humans performing a task together will be the next level of innovation that organizations will have to contend. But the transition into such a situation will neither be easy, nor straightforward. Organizations will have to strengthen their fundamentals. The AS-IS process knowledge should be accurately documented and should be in sync with what exists. Any discrepancy in the AS-IS process understanding will only magnify the rework requirements.
- Make integration of AI and RPA more seamless: Today, we have an assortment of technologies to stitch AI and RPA together. But the patchwork is visible. Just as the development environment has to be integrated with BOTs, AI and RPA need to coexist more seamlessly. The AI technologies required for BOT creation are reasonably well understood. A framework of AI that incorporates RPA capabilities or vice versa will be the next significant progress in RPA innovation.
RPA will continue to belie promises and underperform as long as these yawning gaps remain. We probably need the next big visionary who has the bold imagination, courage and creativity to bring all three aspects of RPA under one umbrella.