Intelligent automation (IA)
In essence, intelligent automation “takes the robot out of the human”. At its core, IA is a concept that leverages a new generation of software-based automation that combines fundamental process redesign with robotic process automation (RPA) and machine learning. This automation is achieved by mimicking human work activities like language, vision, thinking and learning, and execution. The objective is for automation to mimic the activities carried out by humans and, over time, learn to perform them even better. With a redesigned automated process and minimal human intervention, organisations have the capability to deliver greater business outcomes at lower cost while ultimately improving customer and employee satisfaction. It is crucial to bear in mind that while organisations can use RPA to deliver quick solutions that provide excellent short-term benefits, intelligent automation technologies offer long-term value for businesses, employees, and customers. Many vendors and consultancies are still pushing RPA solutions to support key customer journeys and processes, however we believe that RPA is not the silver bullet our clients once hoped for; due to the reliance on structured data it can be almost impossible to automate end-to-end core processes. Intelligent automation offers a more pragmatic alternative to drive more efficient, effective processes. Organisations that have already deployed RPA to perform high-volume, repetitive tasks, should move beyond and re-imagine processes for RPA to mature into intelligent automation.
Figure 1 An orchestrated use of multiple technologies within a process moves an organisation from simple task automation to intelligent automation
Intelligent automation technologies automate human activities by mimicking four capabilities:
- Language (using language to read, speak and write)
- Vision (viewing the world around us, recognising objects)
- Thinking and learning (making decisions, analysing data, creating insights, predicting), and
- Execution (clicking, typing, filling in data)
To successfully mimic these capabilities, IA leverages these core technologies:
- Robotic process automation: software automation technology to automate routine tasks – the tasks that can be done by codifying rules. The robot has a user ID and performs rule-based tasks like humans do, accessing applications from the front end. Key players in the market currently are UiPath, Blue Prism and Automation Anywhere
- Smart workflow: process-management software tools that orchestrates actions across users, robots and systems, allowing users to initiate and track the status of an end-to-end process. They are also referred to as intelligent business process management (iBPM) and low-code tools. Examples of vendors include Appian, AgilePoint and Pegasystems
- Machine learning (ML): whilst there are many definitions, machine learning usually refers to the changes in systems that perform tasks associated with artificial intelligence. Such tasks involve recognition, diagnosis, planning, prediction etc2
- Natural language processing (NLP): the branch of artificial intelligence concerned with giving computers the ability to understand text and spoken words in a similar way humans can. NLP combines rule-based modelling of languages with ML to enable this technology to process language in the form of text and ‘understand’ its full meaning, including the writer’s intent or sentiment3
- Optical character recognition (OCR): technology used to convert virtually any kind of image containing written text (typed, handwritten, or printed) into machine-readable text data. Advanced methods like Zonal OCR can be used to automate complex document-based workflows, and
- Cognitive agents: technologies that combine ML and NLP to create a virtual agent capable of executing tasks, communicating, and making decisions.
Figure 2 The four IA capabilities and their associated technologies
By combining RPA with other technologies such as ML, NLP, smart workflows and OCR, IA goes far beyond simple task automation. It empowers organisations to reduce or eliminate manual efforts and streamline workflows to deliver visible process improvements. As organisations gain knowledge and experience with implementing automation solutions, they will start to combine these technologies in their processes. By orchestrating use of multiple technologies and re-imagining business processes, organisations can move beyond simple task automation and mature into IA. This can help them deliver stronger business results at lower cost, while improving customer and employee satisfaction and generating greater profit per employee.
2 N. J. Nilsson, in Introduction to Machine Learning, 1998, pp. 10-12.
3 I. C. Education, “Natural Language Processing,” 2 July 2020.