Business Translation - AI, Automation, & Robotics
It is understandable that executives are scratching their heads when it comes to understanding or differentiating between AI, automation & robotics. These logic-based capabilities have a lot to offer businesses, especially in this period of digital transformation. Below is some high-level information on these terms.
Artificial Intelligence (AI), Automation and Robotics are synonymous in that they require input(s), execute some logic/calculations and provide an output(s). At a high level the differences are the type, timing and number of inputs and outputs, the complexity of the calculations and the required computing power. At the end of the day these digitized logic capabilities speed up processes while increasing quality and reliability of data, provide better information, and reduce costs.
Interest in these capabilities has been growing more significantly since mid-2016. The graph below shows the trend of these terms on Google in the U.S. AI has risen more substantially over this period versus automation and robotics. The search interest in robotics and automation is about half that of AI.
AI is a subset of computing and has the characteristics of learning, reasoning and solving problems. Its applications can be simple (narrow) to very complex (broad); moving from insight to foresight. AI's advancement is due to advances in computing and the growth of data sources, such as big data. Big data handles a much higher volume, velocity and variety of data.
But for AI to work to its fullest capacity, it must digest its results that will allow it to learn over time. The new outputs get better due to the learning input. A good example of where AI is being applied is “enterprise” technology companies who create tools for workplace roles and functions for example: BI, productivity, customer management, HR, and sales & marketing, plus many more.
Example: Luminoso uses AI to create visual representations of customer feedback, allowing companies to better understand what consumers want.
Robotics or Robotic Process Automation (RPA) interpret user interfaces, and are configured to execute steps identical to a human user. User interfaces can be auditory or textual such as screens or phones. Software robots are configured (or "trained") versus being programmed using code-based instructions.
Robots are like virtual workers for example a Chatbot. Chatbots conduct conversations via auditory or textual methods. Chatbots can use natural language processing or scan for keywords, then present the most appropriate output. An example of this today is customer service via a chat panel. Users a led to believe they are interacting with a human when in fact they are interacting with a Chatbot.
Example: Blue Prism’s “self service” robotic automation technology platform was used to achieve significant improvements in key processes.
Automation or Business Process Automation (BPA) is defined by Gartner as the automation of complex business processes and functions beyond conventional data manipulation and record-keeping activities, usually through the use of advanced technologies. It focuses on “run the business” as opposed to “count the business” types of automation efforts and often deals with event-driven, mission-critical, core processes.
BPA is not new though digital transformation has put a spotlight on it. BPA is a linear process in that there are inputs, data manipulation or logic application, and then an output. No learning feedback loop. The output could be to a user interface or a sub-process within a larger process. Coding versus learning is the method to automate processes.
Example: Appian was used to provide a more efficient way to manage back office operations for one of its clients, eliminating the physical transportation of forms and files.
In summary, learning is key for both AI and robotics. Learning is fed into AI in a continuous process building upon itself, where learning is coded in robotics. Robotics and automation are focused on replacing workers with digitized processes. AI uses advanced computing power and data handling. And, automation handles complex processes, where robotics is a more simplified computing capability.
All three of these logic based capabilities have a place in digital transformation and will grow in interest and adoption as technologies become more affordable and businesses find appropriate applications. Executives should be evaluating these capabilities as part of their digital transformation.