convergenceof data, process and technology has unlocked a plethora of opportunities forbusinesses to create value, save costs and produce real competitive advantages.What seemed impossible ten years ago is now a reality thanks to data, mostrecently termed the world’s most valuable resource, and rightly so.
Data fueledtechnologies such as Robotic Process Automation (RPA), Intelligent Automation,Cognitive Automation, Artificial Intelligence (AI) and Machine Learning are thelatest buzzwords in the global marketplace. Many of these terms are usedinterchangeably. Are they the same, but different? Someview RPA and AI as a continuum, a journey from one to the other. A contrastingview projects RPA and AI to be two different technologies with differentpurposes nudging organizations to make a choice between them. Nevertheless,it’s clear that intelligent automation is here to stay. RPA is poised forexplosive growth given that every data reliant firm has a significant potentialfor adopting automation.
A recent report on disruptive technologies suggeststhat automation technologies will have a potential economic impact of $ 6.7trillion by 2025. This would evidently pave the way for large scaleimplementations set to take place all over the world.
Everycoin has two sides, and so does RPA. EY’s research indicates that while RPA cantransform the economics and service level of current manual operations, 30 to50% of initial RPA projects fail. A pragmatic view of RPA implementationssuggests that RPA projects might not necessarily be delivering expectedbenefits. RPA could expose the organization to unforeseen risks that are givenlittle thought at the time of implementation.
For instance, a multi-billiondollar American airline suffered losses of more than $150 Million when systemssuffered from misconfigurations. Misconfigurations were pushed out quicklythrough automated systems, bringing the systems down and delaying flightsimpacting customer goodwill on a massive scale. This is a clear cut example ofhow customer experience can be impacted when automation goes wrong. Further, inthe arena of customer service, unmanned bots are not always adept at handlingintricate customer interactions, leading to poor consumer experiences. Cyberrisks arising from interconnected automation technologies are a reality thatought to be dealt with head-on, especially in manufacturing and engineeringindustries. Often, multiple vendors are needed to create an automated warehouseor controlled distribution centres, each vendor bringing with it the risk ofaccess to information. This exposes the organization to new avenues of systembreaches since RPA requires access to company databases, effectively leavingdata open to vendors.
Ransomware attacks on industrial automation systems areincreasingly becoming a menace leading to loss of control over these systems. Forexample, a municipal transportation agencyin the United States was infected with ransomware resulting in itsturnstiles being blocked. The attackers demanded a ransom of almost $100,000 tounlock the infected systems. Due attention should also be paid toregulatory and compliance related risks. RPA solutions could be bought fromcountry A, the process might be based in country B and the customer servicedcould be based in country C.
Complexity of combinations make it even moreimperative that implementations take into account applicable legislation andensure compliance. Determining who is liable for what in case theimplementation fails, is of paramount importance.What can companies do to ensure the sustainedsuccess of RPA programs? A three point framework could potentially steerorganizations in the right direction. Firstly, incorporate the right kind ofgovernance mechanisms to ensure smooth change management, vendor management andbenefit tracking. Secondly, it is crucial that the algorithms built at the veryoutset are well designed and trustworthy.
Errors in this could potentially bethe root-cause for multiple issues causing a down-stream domino effect. Lastly,a continuous testing and monitoring process can ensure that the RPA environmentis audited remotely in real time.Professor Shoshana Zuboff of Harvard BusinessSchool asserts, “Everything that can be automated will be automated”. Successfullymaneuvering the maze of intelligent automation will depend on the ability oforganizations to govern, operate, monitor and improve upon the technology. Apro-active approach to intelligent automation will take companies a long way inminimizing the risks that come with both, large and small scaleimplementations.