Moore’s Law, Speed and Robotic Process Automation


Speed is one of the great benefits touted about robotic process automation – but not enough people are asking about it today.  Beyond cost reduction, most leaders today looking at RPA are focused on things like reduction of errors, better management of repeatable tasks, and process standardization.  Beyond the obvious fact that robots are cheaper than using human labor, the question of “How fast can your robot go?” will soon be a major question.  Speed of process throughput.  Speed to insight and intelligence.  Twice as fast as humans?  Twenty times as fast?  Thirty?  Where’s the limit?

But one of the great advantages of using RPA technologies -- that they can be overlaid onto existing systems – is also one of its hindrances: processes automated by RPA can only go as fast as the capabilities of the legacy systems that underpin them. 

Consider this “speed impediment” as a kind of variant of Moore’s Law applied to RPA.  Moore’s Law, of course, posits that the number of transistors per square inch on integrated circuits had doubled every year since the integrated circuit was invented.  Beyond today’s fascination with the newness of RPA (and yes, accelerating a process to 20-30 times as fast as a human is pretty cool), at some point in not-too-distant future, businesses are going to wonder where the next quantum leap in speed of “speed of process processing” will come from.  Consider the need for precise, to-the-nanosecond execution of high-speed trading in banks, for example.  If the underlying core systems creaky and “otherwise-so-old-they-could-walk-with-a-cane”, it will limit the effectiveness of RPA to drive the advantages speed, let alone systems maintenance costs and throughput volume.   

What’s all this mean for Moore’s Law and the future speed of robots in enterprise processes?

A future development to help “break the sound barrier” of robotic speed may look something like Ray Kurzweil's law of accelerating returns (LOAR), in which a scenario of “self-replicating robots that make faster bots” comes to pass.  But rather than get too spooky with possibilities or technological dystopia of The Singularity, more likely, the future is one that’s digital, using next-gen SMAC technologies like mobile platforms, sensors, telematics, and the internet of things where processes are “born automated” – and born lightning fast -- from the outset. 

It’s very likely RPA is only the first overture of an epic drama about to play out and change enterprise processes as we know them. And the REAL main event of Act One will be speed and intelligence that comes about through digital process transformation.  It’s not about “robots”, per se, but about the complete retrofit or creation of digital processes.  In fact, there may be a gnawing concern that simply applying a robot to an “as-is” process, still leaves organizations woefully short of the truly differentiated ballet of today’s high-flying, competitive outliers that have disrupted entire industries through process digitization.

For example, think about how easy (and fast) paying for a bagel is with a digital BPaaS like Square or Apple Pay.   Imagine if those technologies did not exist, and all we had was a “robot” that replaced a human cashier pulling the handle on an old-school, “dumb” point-of-sale cash register. The robot could be lightning-fast, but the cash register (the underlying system) would be painfully slow – or worse, analog to the point of only using paper-based cash and receipts.  Mary Meeker at Kleiner Perkins recently put it this way: “While consumer Internet entrepreneurs often pursue personal passion... Enterprise Internet entrepreneurs often pursue prior company pain points”.  Aaron Levie, CEO of Box, had this to say: “Adding software to a broken process doesn’t make you digital.  The biggest challenge is reimagining the process, not writing the software”.

Process digitization can also radically accelerate and transform data analytics — and business models. Our new research among 537 leaders shows that nearly one-third cite improved quality/consistency/believability in the data they’re getting from digital process initiatives, and 27% report easier integration across processes. What’s also changing is the impact of digital processes on value chains and operating models. Each industry and its processes — whether claims management in insurance, or reconciliation or mortgage processing in banks — is swiftly adopting new process models. Harnessing the power of digital smart machines and AI, relationships that were traditionally transactional are now “interactional;” that is, rather than being “once-and-done,” they involve multiple interactions, and the more you can learn about a customer, supplier, partner — or even employee — the more meaningful each subsequent transaction can become.

Because of these forces, true digital transformation of processes will likely start to take off at an exponential rate, spurring businesses a way to break from the past of legacy systems, processes, and procedures to enable exponential growth that outpaces that of rivals that won’t (or can’t).  Speed of process is going to play an important role, and with Moore’s Law as a guide and an ability to re-imagine the nature of the process and the work, the future possibilities – and the notion of a “speed limit” for automated processes – will be endless.