operators working together to decrease process time

Here's an example of using BasicFRs as processors to have more control over processes. The BasicFRs call the idle operators who are most skilled at this process to help (up to 20, which is set in a label), it sums up their skills and divides the process time by this sum. I wrote and commented custom code in the BasicFRs and the Dispatchers' Send To trigger. The BasicFRs are identical except for their labels, and the dispatchers are identical. I've also configured a multiple table excel import to get the data from excel. It is possible to write Post Import Code to make, name, and connect operators based on the table, but I didn't want to make this model that complicated.

I like your idea of a sample model showing how to have more control over the operators. The model's fun to watch too as the mass of operators scramble around the model looking like a bunch of crazed ants.

As another example, it would be good to give everyone a sample model that does it all:

1. allow multiple operators up to a maximum
2. screen operators based on operator attributes
3. adjust the process time based on how many operators show up (not just who are available at the time the work is started).
4. adjust the process time based on the attributes of the operators who show up.
5. allow the process time to begin after a minimum number of operators show up, and then be recalculated as more operators arrive.
6. allow operators to be preempted away from a job, and so long as there are a minimum number of operators still on the job, let the job continue but at a slower pace.

I'm addressing these issues with the new healthcare library, and they are quite challenging though doable. I'm using coordinated tasksequences with subtasks. One thing that would make it all easier is if allocations could be cancelled...

Anyways, I thought I would give you a challenge to keep you from getting too bored as you invent the next Rubiks Cube during your lunch breaks

Here's what Cliff is talking about. I made a Rubik's Cube object in Flexsim. There's a user event sending it random messages 1-18 every pi/2 time units for the 18 different rotations a Rubik's Cube can do. It's just one of those things to show how cool Flexsim is. If you can model a Rubik's cube, you can model anything.

Because Alex' simulation model of a Rubiks Cube requires compiling (I guess flexscript was too slow to crunch away at his multi-dimensional arrays contained in tree format, so he switched to C++), here's a video for those of you who don't have and don't want to install Microsoft Visual C++ 2008 Express Edition (even if it is free).

Anyone who reads this and doesn't already have it, please go to http://www.microsoft.com/express/download/ and download Microsoft Visual C++ Express 2008. It's free, and even if you don't have Flexsim, I recommend having it. It takes about an hour to download and install, but it's a powerful program.

Flexscript is fast enough for this model. I used C++ because I used an OpenGL command (multmatrixd) that isn't implemented in Flexscript.

I'm not sure I like the caption on your video, Cliff.