Master´s thesis addresses automation of knowledge work

​The purpose of the project conducted during year 2016 was to identify and explore opportunities behind automation of knowledge work. As a result of research, a service concept was created around co-creation of content behind knowledge work automation. The concept can be used either as a company-internal or cross-company service. It enables organizations to bring their physical and virtual workforce (aka software robots) into an environment where they can all learn from each other. At the same time it enables the knowledge worker to remain in control of own skills and proactively develop them.

Knowledge work and automation content

Knowledge work is today mostly understood as work which involves dealing with information to perform complex analyzes, make subtle judgments and do creative problem solving. Increasing parts of this kind of work can be outsourced to the machines.
Some typical examples are collecting information about clients credit rating, making decisions on whether to grant a housing loan or accept an insurance claim, booking meetings, or even performing first line customer support. This is often perceived as a threat taking away knowledge workers’ jobs. 

The automation of these kind of tasks is enabled by development of general software technologies, with robotic process automation and artificial intelligence in particular. The work is performed automatically by creating automation content and feeding that content to a robot. The automation content are instructions and manuals based on which a robot can perform a business task, independently or partially assisted by a human.

Bi-directional and reusable

Automation content, once created, typically becomes part of a software tool given to knowledge workers to do their daily jobs. The knowledge put into the content is thereby fed back to the ultimate source of that knowledge, i.e. knowledge worker. In that sense, the content acts as a bi-directional link between knowledge worker and a robot.

However, the research shows that the user often considers software tools as being organization-imposed siloes of information, requiring many repetitive user actions and often not being up to the task. This indicates that the robot-worker direction does not always work well.

The research also shows that it is important for the worker to learn from the peers and know the reasoning behind different ways of working. This indicates the importance of reusability of knowledge, also the one built into the automation content. At the same time it was recognized that every organization and person have own ways of working which might limit reusability. Atomization of automation content, when presented to the users as one possible solution, was perceived as a possible solution to bridge this gap.

Inclusive communities

Creation of the automation content still remains domain of specialists. The typical creation process involves pre-coding the business rules with the help of business domain expert. 

Another way is helping robot learn the business rules by feeding it with data created by the domain expert. Thereby the content creation community remains open only for the persons with software development and data science skills.  Notable examples of such communities are e.g. Kaggle and Algorithmia.

However, the research reveals that there is a category of knowledge workers who would be willing to create automation content themselves, if it was simple enough and the benefits would be obvious. This indicates the need for an inclusive community, where the domain experts would be able to not only use but also create content themselves.

Co-creation of automation content

In his thesis project in Laurea’s Degree Program in Service Innovation and Design Mr. Nenad Veselinovic proposed a framework to realize the above identified opportunities. The concept was validated for feasibility and usefulness by interviewing ten management representatives of a typical knowledge work organization and an automation service provider. The concept was also validated for value creation potential by means of business simulations.

More information:

Nenad Veselinovic
Degree Program In Service Innovation and Design
Tel +358 40 779 5483

Modified 2/28/2017 3:24 PM