Levels of automation - a triology (part 1)

The simplest and perhaps most common automation that comes to mind when talking about automation is the robot, but there are many different types of solutions that help a person to relieve both body and mind.

Within an assembly station there are many different tasks to be performed, these tasks can be performed by different resources that take the help of different hardware and software.

The more help a person takes from the different types of hardware and software the higher the degree of automation they use.

Physical automation is when the body gets help, this can for example be that you get help with carrying, placing, screwing etc

Cognitive automation is when the mind gets help, this can be, for example, remembering where screws should be located, which screws should be used and when to do what.

Some companies talk about manual and automatic stations others add semi-automatic stations. In order to be able to make as good a solution for automation as possible, companies need to break down the station into smaller components, such as tasks. On the one hand, they need to look at solutions that include help for humans in several different ways, this gives a need to measure several different degrees of automation.

With this starting point, Chalmers, together with companies and other researchers, created a strategy and method for choosing different types of automation solutions depending on what the task looks like. Below are some examples of physical and cognitive automation. In part 2, you can read about the seven levels used in automation selection.

By combining help for both body and mind, advanced tasks can be made easier to perform.

Examples of levels of physical automation

An example of a task can be screw a screw, this can be done in different ways;

- By hand

- With a simple screwdriver

- With a battery-powered (or air-powered) screwdriver or

- With a robot

All these different solutions are examples of different degrees of physical automation that help people to screw in a screw, the task is the same but the solutions are different. Different solutions also result in different results.

A too low degree of automation can, for example, make it difficult to screw a screw with the right torque, it can also take time if more screws are to be screwed in (everyone who has built the terrace may be able to relate to this).

Examples of levels of cognitive automation

We continue with the same example of a task can be screw in a screw, this can be done in different ways;

- With my own experience

- With a simple description that describes the type of screw to be used

- With an instruction (or code) that describes where the screw should be located and with what force it should be screwed in

- With an instruction (or code) that shows where the screw should be located, with what force it should be screwed in and which tells the task is performed incorrectly

All these different solutions are examples of different degrees of cognitive automation that help people to screw in a screw, the task is the same but the solutions are different.

With higher levels of physical automation, instructions made for humans are exchanged for instructions made for, for example, a robot (ie robot code)

In the same way as for physical automation, too low cognitive automation can result in different results. If instructions are missing, it may, for example, take longer to figure out on your own where a screw should be located.


Combination of physical and cognitive automation

By looking at both the physical and the cognitive part of a task, companies can see which of these needs help. In complicated tasks, a higher degree of help may be needed for the bud, while in heavy lifting, more help is needed for the body. The solution or degree of automation that you choose has different results. This result can be measured in different ways, most often companies calculate this in KPIs (Key Performance Index).
Examples of KPIs can be efficiency, cost and quality. Depending on which of the different solutions is chosen for a particular task, this will have an impact on the KPIs.

If you want to know more about methods that can be used for both division of tasks in a station and selection of automation, you can contact Åsa Fast-Berglund (asa.fasth@chalmers.se) and she can lecture about this.

You can also buy the book Smart automation where these two methods (task allocation and automation strategies) are included as two methods out of seven.


This article is tagged with these tags. Click a tag to see all the articles with this tag.

This article is categorised as Introduction  |  Published 2018-04-03  |  Authored by Åsa Fast-Berglund