Modern workflows, whether in industrial or IT sector, are all about
optimizing the work force and work flows. Automation in the form of Robotic
process automation (RPA) comes as a boon to handle operational tasks with
least manual intervention. However, for the processes that require logical
reasoning and judgement, RPA falls short. And that is where Artificial
Intelligence (AI) comes to the rescue. So far, Artificial Intelligence is
seen as shown in the Sci-Fi as a dangerous alternative which can overcome
human species. But in real world, AI is helping businesses such as Facebook,
search engines, and even your friendly neighborhood emails in optimizing
operations and achieving high level of user happiness.
To intelligently automate end-to-end processes and enhance value beyond
efficient processing, companies are copying up to the idea of the Automation
Continuum, combining systems that "Do" (RPA) and systems that "Think" and
"Learn" (Artificial Intelligence (AI) and Machine Learning (ML)
RPA: Complex yet simple
RPA solutions are limited to handling rule-based work and need digitized and structured inputs. The RPA robots do exactly what you tell them to do, and they will do the same all the time, which is desirable if you want to achieve compliance and accuracy. RPA focusses on the human tasks and works on the User Interface and can be instructed to write data in a set pattern in continuity.
AI: Putting mind to work
AI provides the ease of judgement-based processing and unstructured inputs. Where there is any ambiguity, usually when the inputs into a process are unstructured, such as emails, or where there are very large amounts of data, then AI is the appropriate technology to use because it can 'understand and manage' that variability and, most importantly, learn the patterns over a period of time, for example, in handling free-text emails or invoices. AI is excellent self-learner, that is, over a period of time, it can analyze the information pattern and provide options and answers. For example, google mail learns to identify spam from the regular emails by analyzing the patterns and key words of the spam mails. The AI technologies are excellent in capturing information such as:
Vision Recognition: recognizing a face in a photo
Sound Recognition: transcribing words that someone is saying Search: extracting data from unstructured or semi-structured documents
Data Analysis: identifying patterns of behaviors in customer data
three of these operations require Supervised Learning, that is, they require large data sets to learn the necessary patterns, whereas the data analysis uses Unsupervised Learning, which means that it can come up with the answers without you telling it what the question is. Therefore, AI turns the unstructured data into information.
RPA or AI: A choice, not an alternative
Whether to use RPA or AI, or both together is just a set of choices based on the specific demands. The two technologies complement each other very well; for example, by using AI to structure the unstructured data at the beginning of the process, and then by using the robots to process the transactions, and at the end using AI for decision making and data analytics.