Skip to main content

I would like to fill polygons with Circles of a designated size. Basically show me how many full circles will fit inside each polygon.

 

This is of course related to figuring out a maximum number of people could physically fit within the various areas of our airport and maintain a certain level of physical distancing. I could figure out some methods using the Tiler transformer but these would be squares and will impact the results.

 

Here is a link to someone else who was trying to figure out the same issue.

https://gis.stackexchange.com/questions/193482/how-to-calculate-how-many-polygons-i-can-put-inside-a-polygon

The custom transformer 'HexagonSampler' will generate hexagons of a specific size to fill your polygon


Thanks for the quick answer. The 'HexagonSampler" is closer but it simply creates a pattern of Hexagon's similar to the "Tiler"'s Squares. It does not maximize the number of full Hexagon's that can fit in the polygon like I would like to do with Circles.

As you can see in the link I provided I am looking for results that are not necessarily a pattern.


You were right with the tiler. Tile the polygon, center point replace, and then buffer.

 

Attached fmwt

 

Polygon

circles1Tiled

circles2Center point

circles3Buffered circles

circles4


Better get studying circle packing algorithms!

 

Realistically, I'm not sure that the absolute maximum number of people you can get distributed with a given distance between each is the figure you want to "figuring out a maximum number of people could physically fit within the various areas of our airport and maintain a certain level of physical distancing"

 

I would have thought the hexagon tiler would be a good starting point, although even then, assuming the people in your airport are moving it would require much more sophisticated analysis to work out a sensible maximum.

 


Thanks jlbaker2779 for the sample workspace. It demonstrates the circle result from the Tiler could definitely work.

Thanks ebygomm for pointing out this was called "Circle Packing". The reference material I found on it quickly surpassed my knowledge of geometry, but showed that fairly solid algorithms exist for putting this place. Nothing I found looked like it would "Plug" into FME but I may be wrong.

 

As far ebygomm's comments on serious limitations of trying to determine a theoretical limit of people in the airport I completely agree. There are much better and costly ways to do this sort of analysis that will give much more meaningful results. We are just looking for something better than "the public area your airport is Xm² therefore you are only allowed this many people in the terminal with social distancing in place". For planning purposes, if things start to ramp up again and some variation of social distancing is required, we need to figure out some way of determining at what passenger count we are going to start running into issues.


Thanks jlbaker2779 for the sample workspace. It demonstrates the circle result from the Tiler could definitely work.

Thanks ebygomm for pointing out this was called "Circle Packing". The reference material I found on it quickly surpassed my knowledge of geometry, but showed that fairly solid algorithms exist for putting this place. Nothing I found looked like it would "Plug" into FME but I may be wrong.

 

As far ebygomm's comments on serious limitations of trying to determine a theoretical limit of people in the airport I completely agree. There are much better and costly ways to do this sort of analysis that will give much more meaningful results. We are just looking for something better than "the public area your airport is Xm² therefore you are only allowed this many people in the terminal with social distancing in place". For planning purposes, if things start to ramp up again and some variation of social distancing is required, we need to figure out some way of determining at what passenger count we are going to start running into issues.

@Tim Albert​  Looks like the hexagon pattern does give the tightest circle packing. Have a look at http://www.packomania.com/ !


Thanks jlbaker2779 for the sample workspace. It demonstrates the circle result from the Tiler could definitely work.

Thanks ebygomm for pointing out this was called "Circle Packing". The reference material I found on it quickly surpassed my knowledge of geometry, but showed that fairly solid algorithms exist for putting this place. Nothing I found looked like it would "Plug" into FME but I may be wrong.

 

As far ebygomm's comments on serious limitations of trying to determine a theoretical limit of people in the airport I completely agree. There are much better and costly ways to do this sort of analysis that will give much more meaningful results. We are just looking for something better than "the public area your airport is Xm² therefore you are only allowed this many people in the terminal with social distancing in place". For planning purposes, if things start to ramp up again and some variation of social distancing is required, we need to figure out some way of determining at what passenger count we are going to start running into issues.

Everyone needs to space out in the airport so they can sit next to each other in a tube. :/

 

Our job is not to ask why. Our job to make the maps.

 

You're probably better off creating a map of risk zones vs. spacing. Create a set of maps with red/yellow/green zones, plaster them everywhere, and put stickers all over the place on the walls and floors. That's a pretty cheap way to accomplish spacing warnings.

 

The solution is dilution. Having a limit on the number of people in an area won't make much of a difference if they aren't aware the area is a high risk area and then cluster up.


Reply