What Is Cognitive Automation?
Cognitive automation, also known as intelligent automation or smart automation, can efficiently process high volumes of both structured and unstructured data. It uses Ai technologies including machine learning, computer vision, Natural Language Processing and fuzzy logic to organise the data, analyse it and draw conclusions.


Reading time

3 minutes
October 7, 2019

What Is Cognitive Automation?

What Is Cognitive Automation?

Cognitive automation, also known as intelligent automation or smart automation, is software that can be used to extend the functionality of Robotic Process Automation. While RPA can efficiently process high volumes of text-based, structured data, it is unable to work with unstructured data that does not fit within a pre-defined model. Cognitive automation works to process this unstructured data through Ai technologies like machine learning, computer vision, Natural Language Processing and fuzzy logic. It then organises the data into a structured pattern so it can be analysed and draws conclusions from that analysis.

What can Cognitive Automation do?

This technology’s strength lies in its ability to adapt to new types of problems, build knowledge, understand human language and overcome ambiguity. Cognitive automation’s ability to process both structured and unstructured data allows organisations to automate processes from end to end. It can also be used to link traditional RPA software to pure cognitive platforms like Microsoft Cognitive Services and Google Cloud AI. We have put together some examples of how it can drive efficiency at work.

Human language processing

In any organisation, there is a large quantity of unstructured data that can be processed by cognitive automation and translated into structured digital documentation. This data includes spoken language, handwritten files and digital files that do not follow a specific format.

Examples of application:

  • Transcription of phone calls for inclusion in client files
  • Transcription of meetings
  • Recording surgeons commentary throughout surgery and recording within a patient’s record
  • “Reading” handwritten patient files and translating them to digital files
  • Interpreting invoices to extract the invoice number, supplier name, due date and other pertinent information


Image Processing

Cognitive automation uses computer vision and machine learning to interpret images and make decisions based on the findings. It has taken some time for the technology to reach its current maturity due to the inherent complexity of the visual world. However, there are several crucial advantages to computer image processing. Firstly, computers are to process and retain limitless amounts of information. So while a radiologist may have seen 10,000 CT scans over their career, a computer can analyse and compare millions of scans and retain this information and learning going forward. And the same information can be programmed into every radiology computer programme in the future. Secondly, a computer doesn’t make mistakes and can work 24 hours a day!

Examples of application:

  • Analysis of CT scans and other medical imagery reports and diagnose diseases
  • Inspection of products for non-conformance or defects 
  • Use of pattern recognition to detect irregularities in data sets


RPA analysis, optimisation and orchestration

Like any software, RPA has a wide range of features and functionality. For an organisation to get the most value from the software, it needs to be appropriately integrated into the companies processes and then regularly reviewed. Reviewing RPA bot performance to optimise within the business is time-consuming and requires specialised knowledge within the organisation. Additionally, any single organisation will typically have hundreds of bots deployed at any one time. Through the use of ML algorithms, the orchestration of these bots can be automated.

Examples of application:

  • Identifying any bottlenecks in the process
  • Allocating bots to specific tasks within the business
  • Deploying new bots
  • Scaling the number of working bots


Error handling

Computer software is very precise in its actions. This precision has an enormous amount of value in an industry like healthcare, where a decimal point error in prescription medication could be fatal. Cognitive automation can be used to flag potential mistakes for human review. This functionality is also beneficial when it comes to compliance with regulatory requirements, health insurance claims and account reconciliation.

Examples of application:

  • Verifying information for claims processing
  • Identifying data discrepancies
  • Revealing duplicate entries


Cognitive automation is instrumental in driving efficiency in healthcare professionals’ workflows, improving patient experience and having a positive impact on the organisation’s bottom line. But most importantly, it helps free up staff to focus on value-adding work.