This post is the first post in an upcoming series of posts on AnyLogic simulation modelling. I want to introduce basic material handling system simulation modelling using AnyLogic as a discrete-event simulation machine and modelling environment. I will use AnyLogic’s material handling library for the examples in this post, introducing some simple conveyor-centered DES models.
To start with I introduce a simple conveyor line model with one simple roller conveyor
Below tutorial introduces conveyor space markups and the conveyor block in AnyLogic. It also makes use of the process modelling library in AnyLogic for modelling a source and a sink.
The source creates material item agents. You can create custom material item agents – as introduced in the video. E.g. I can customize a material item agent type to represent boxes to be moved along the conveyor.
Next, we can add a spur to the conveyor line to model conveyor spurs
In below tutorial the previous model is extended to comprise a conveyor spur space markup, modelling a conveyor spur that separates the initial conveyor line into two separate conveyor lines. Material flow at the spur can be controlled with a “selectOutput” modelling component from the process modelling library in AnyLogic. This is demonstrated in below video.
I proceed by adding a work station to one of the conveyor lines
A work station allows for modelling a process associated with the conveyor line, e.g. an assembly process. The work station allows for specifying e.g. processing times and their distributions. This is demonstrated in below tutorial.
So far, no ressource was assigned to the work station; this is done now
Ressources and ressource pools allow for modelling e.g. assembly workers that conduct assembly tasks at, in this case, a specified work station. Ressource pools can also e.g. be linked with shift calendars which allows for shift calendar modelling.
Processes can have a defined batch size and also this can be modelled
In this example I model a batch process in which multiple parts are converted into a single part. This could e.g. be a packaging process in which items are packaged into packaging materials such as e.g. carton boxes.
Next I model a mobile assembly workers that are assigned to multiple work stations
Ressources must not be assigned to a static location in the model. They can be mobile and this can be important to model in material handling systems. In this case a worker pool is assigned to three conveyor work stations and the path that workers can take when moving from station to station is defined by a node network.
In upcoming posts I will introduce the remaining components of the AnyLogic material handling library.
Data scientist focusing on simulation, optimization and modeling in R, SQL, VBA and Python