The Robots Are Here

“Any task that we have that can be measured in terms of productivity is a (function) that will go over to the bots, that will go over to the AIs. Anything where you’re measuring efficiency is going to go over. Efficiency is for robots. Efficiency is for AIs.” — Kevin Kelly, The Inevitable: Understanding the Twelve Technological Forces That Will Shape Our Future

Robotic Process Automation, or “RPA,” is commonly understood to mean automation software that can be applied to various commercial uses; RPA software automates tasks that are often done by humans today.

Contrary to what the term itself might suggest, RPA “robots” are actually software programs — and not physical robots — created with varying degrees of artificial intelligence and cognitive computing. They range from relatively simple programs that automate manual and repetitive tasks (that are highly predictable), to advanced algorithms based on machine learning that are able to automate a variety of sophisticated tasks and handle complex exception cases.

While RPA is a subset of the economy-wide AI and automation phenomenon, its typical goal is not to automate jobs wholesale. Instead, RPA seeks to automate tasks within jobs, especially those that are the least interesting to perform by a human. In this way, a major objective for a given RPA implementation is to improve the quality of work within a specific job, enabling a worker to focus more on knowledge work and less on repetitive / low value added tasks.

A recent McKinsey study summarizes the factors that drive the decision to implement RPA: 1) technical feasibility; 2) the cost of developing and deploying automation; 3) the cost of labor and related supply-and-demand dynamics (“if workers are in abundant supply and significantly less expensive than automation, this could be a decisive argument against it”); and 4) the benefits beyond labor substitution, “including higher levels of output, better quality, and fewer errors. These are often larger than those of reducing labor costs.”

Our experience at RapidRPA confirms these factors, especially the fourth one: productivity effectiveness improvements generally beat cost savings — which are significant themselves — in most if not all RPA implementations.

RPA Today and Over the Next Several Years

At the current level of technological sophistication, the capabilities of RPA generally outstrip the implementations… meaning, there is a lot more automation that can be achieved today with current technology than has actually been implemented cross industries to date. In the McKinsey Quarterly article about the study quoted above, the authors note that “(by using) currently demonstrated technologies, (companies) could automate 45 percent of the activities people are paid to perform, and about 60 percent of all occupations could see 30 percent or more of their constituent activities automated.”

And that is using only the technology available today! Meanwhile, advances in the field are proceeding at a staggering pace, driven especially by advances in deep machine learning. We are truly in the first inning of this “smart automation” wave.

Current RPA implementations — which are substantially increasing in 2016 — are most common in larger enterprises. Firms are piloting RPA typically in select back-office environments, particularly shared services operations such as finance, call centers, human resources and so on. RPA is augmenting “swivel chair environments,” shorthand for activities that are routine, repetitive and do not require a great deal of thought or decision-making.

Eighteen Months

Enterprise organizations will begin to push the envelope on RPA where the need for AI capabilities will increasingly begin to unfold. Some machine learning capabilities and “if this, then that” decisioning will be performed by a software-based robot. You’ll also see a greater advance toward middle-office and even some front-office, client-facing activities where the customer won’t even know they’re interacting with a robot.

Most firms will be advancing beyond simple pilots and proofs of concept into full-blown production, and the majority will employ multiple RPA vendor solutions. The need to optimize robot performance, track results and process improvements will be a necessity to report up to management, Board of Directors, regulators and shareholders.

Longer Term

RPA will continue to evolve and augment (if not disrupt) every aspect of business operations. Here are a few predictions based on our experience implementing RPA throughout large enterprises:

  • Industry analysts will be asking firms about their RPA strategy to improve margins
  • RPA centers of excellence will be commonplace
  • Regulators will delve into operations not leveraging RPA fully, and ask “Why not?”
  • Outsourcing and offshoring vendors will have major service offerings centered on RPA as a means to stay relevant with clients; as a result, more insourcing deals will occur
  • The role of human resources professionals will change to include RPA in talent management, objective setting, and training (particularly training staff to handle process exceptions and robot configuration)
  • Back-office roles will climb toward a 50%+ RPA-enabled
  • Emergence of the Chief Automation/RPA Officer role will be common at the most advance firms
  • Increasing reliance on AI solutions as a means of taking RPA to the next level