Artificial intelligence is being applied to a broad range of applications from self-driving vehicles to predictive maintenance. Some of the more mundane, and even boring, applications are focused on helping improve automation of back office operations. There has been a real acceleration in the use of automation tools for back office operation, with much attention (and money) flowing to Robotic Process Automation (RPA) tools. However, these tools are not intelligent, per se. None of the vendors in the space are claiming that their tools can be classified as AI, however, increasingly end-users are seeing process automation tools as solving issues of digitization of processes that are required to adopt more intelligent, machine learning-based approaches for process discovery, definition, orchestration and optimization. It is these higher level, machine-learning based approaches for dealing with these issues that are the beginnings of intelligent process automation, or what some are calling cognitive automation.
The Confusion Around “Bots”
Part of the reason many confuse RPA with AI is the use of the term “robotic”. To the uninformed, they believe robots have aspects of cognition and intelligence. However, nothing can be farther from the truth. Industrial robots have been used on factory floors for decades. These big machines are dumb. They aren’t aware of their surroundings, and they must be explicitly programmed for the specific operations they accomplish. They move in very specific ways and can’t tell if a human or other object is in the way. This is why industrial robots must be isolated from people in production.
Similarly, software robots aren’t necessarily intelligent. Indeed, automation is not intelligence. So what exactly are these bots doing? Their primary function is to repeat and execute basic tasks that humans usually perform. These functions are usually performed within the software or the computer, and they replace more mundane tasks that are performed on the computer by humans. The main difference between some of the other bots we've looked at and software bots is that these are not chatbots, and they are not intelligent. They do mundane but necessary tasks. The benefit of using them is because without them, humans use what we call a swivel chair integration. The tasks they would perform use human workers or virtual assistants to get stuff done. These methods are error-prone, inefficient and cost a lot of money.
RPA solves three key problems. The first is that it replaces the cutting and pasting of information from one place to another. Second, it automates the entry of information into systems. Third, it speeds up or makes these repetitive tasks more accurate across a wide range of systems that can only be interacted with through their web or other user interfaces.
However, repeating the same tasks over and over, while valuable, does not require cognitive technology, machine learning, or anything within the spectrum of AI. RPA bots, like their factory brethren, are good at executing a process, but not making judgment calls. They cannot adapt to changes in user interfaces, data or apps structures. They can't figure out what to do if information that they need is bad, missing, or incomplete. Rather, to be considered intelligent requires at least a modicum of learning. Learning is gathered from experience and the power of machine learning is improving performance over time with that experience. This is not something that rote repetitive operation software bots or current RPA tools.
Moving to Cognitive Automation
Recently, Cognilytica published a report focusing on Intelligent Process Automation. (Disclosure: I am a principal analyst at Cognilytica). The report outlines the concept that the eventual end goal of AI-enabled content automation is to realize the vision of Autonomous Business Process. Just like their self-driving cousins, autonomous business process (ABP) is the concept of processes that can autonomously discover process flows, autonomously optimize those flows and autonomously handle process exceptions and changes. Indeed, just like how we can have autonomous, self-driving vehicles that take humans from point A to point B, so too can we have autonomous, self-driving processes that can take organizations from point A to point B.
The Four Levels of Cognitive Automation COGNILYTICA
According to the report, just like there are six levels of autonomy for autonomous vehicles, there are four levels of autonomy for cognitive automation. At the very lowest level (Level 0), there’s no intelligence or autonomy. This is where RPA tools currently reside. At Level 1, there’s enhanced intelligence in the form of context and user interface awareness. This is usually accomplished through the use of natural language processing and image recognition tools. At level 2, there’s greater awareness of the processes themselves, autonomously handling process exceptions, autonomously documenting processes and dealing with finding patterns and commonalities between multiple business processes. At the highest level of autonomy, Level 3, we have full autonomous business process, encapsulating all the capabilities discussed above.
These intelligent bots have more power than their dumber, repetitive alternatives. Many repetitive processes that often change can be operated without requiring continuous, and expensive, service and maintenance. These intelligent systems can transfer and transform data between different systems. With more intelligence comes more transformative power, giving enterprises the benefit of systems that can respond agilely to changes in environment with more speed than before.
The market for intelligent tools is currently very nascent, with the bulk of vendors providing tools at Level 0 and Level 1 of Cognitive Automation. According to the report, this market is growing from eight hundred million dollars in 2017 to 8.3 billion dollars in 2023. However, by 2023, these tools will gain significant capabilities with intelligence and machine learning. Will we see any vendor providing tools at Level 3 capability? Just like with autonomous vehicles, that remains to be seen, but the race is on, and we’re hopeful to see the truly transformative power of cognitive automation tools.
Ronald Schmelzer, columnist, is senior analyst and founder of the Artificial Intelligence-focused analyst and advisory firm Cognilytica, and is also the host of the AI Today podcast, SXSW Innovation Awards Judge, founder and operator of TechBreakfast demo format events, and an expert in AI, Machine Learning, Enterprise Architecture, venture capital, startup and entrepreneurial ecosystems, and more. Prior to founding Cognilytica, Ron founded and ran ZapThink, an industry analyst firm focused on Service-Oriented Architecture (SOA), Cloud Computing, Web Services, XML, & Enterprise Architecture, which was acquired by Dovel Technologies in August 2011.
Ron is a Parallel Entrepreneur, having started and sold a number of successful companies. The companies Ron has started and run have collectively employed hundreds of people, raised over $60M in Venture funding and exits in the millions. Ron was founder and chief organizer of TechBreakfast – the largest monthly morning tech meetup in the nation with over 50,000 members and 3000+ attendees at the monthly events across the US including Baltimore, DC, NY, Boston, Austin, Silicon Valley, Philadelphia, Raleigh and more.
He was also founder and CEO at Bizelo, a SaaS company focused on small business apps, and was Founder and CTO of ChannelWave, an enterprise software company which raised $60M+ in VC funding and subsequently acquired by Click Commerce, a publicly traded company. Ron founded and was CEO of VirtuMall and VirtuFlex from 1994-1998, and hired the CEO before it merged with ChannelWave.
Ron is a well-known expert in IT, Software-as-a-Service (SaaS), XML, Web Services, and Service-Oriented Architecture (SOA). He is well regarded as a startup marketing & sales adviser, and is currently mentor & investor in the TechStars seed stage investment program, where he has been involved since 2009. In addition, he is a judge of SXSW Interactive Awards and served on standards bodies such as RosettaNet, UDDI, and ebXML.
Ron is the lead author of XML And Web Services Unleashed (SAMS 2002) and co-author of Service-Orient or Be Doomed (Wiley 2006) with Jason Bloomberg. Ron received a B.S. degree in Computer Science and Engineering from Massachusetts Institute of Technology (MIT) and MBA from Johns Hopkins University.