So, aimed to finish the isovist off to provide us a rough indication of how many Raspberry Pi’s we may need for each environment and extracting the xy coordinates to be able to start visualising information.

The script is starting from Alex’s existing one living in here. Started by placing some sample points throughout each office. The computing part takes a while depending on how many points there are, so I did a rough test by using those points and adding circles with a radius of 8000m and guessing by eye where the the isovist will start to take shape.

Made sure that there is some overlap in the radius to make sure that all areas are covered. The radius of the isovist is at 8m as the results from the last experiment seen here can still be detected. Can also assume that results start to get further lower by 10m. Although just still to 8m for now.

Once I got a rough indication, let it run for a good 30 mins or so for each environment. Also, just an additional note, the furniture is not part of the isovist, only the immediate boundaries like walls, columns etc.

Starting at the BVN office, images say we needed around 27 base stations. Can argue that if we add more, it may mean better accuracy, better results etc.

BVN Base Stations

Based off the isovist script, ARUP will need 34 base stations. ARUP has many nooks and crannies in the office, so again, can also argue for more basestations.

ARUP base stations

Just some things to take note of:

  • If you look closely at any of the images, I closed off individual rooms so that the isovist is bounded by walls, even where the doors are - reason being is that as most of these rooms has the door closed, thought it would be more applicable.

  • This isovist model isn’t a very accurate reading - but rather an educated guess of how the signals would travel in the worst case scenario. In real world situations, the isovist model will not likely to be a perfect radius coming from the last post here. Also, the different environments would react differently to different situations.

The next part was extracing the XY coordinates. This is part of a simulation/animation that BD had started to brew up. Coordinates of BVN and ARUP can be found on github right here.