OneByOne: First map every wall as a line segment with its start and end coordinates. Then create a graph where each segment is a node and edges connect segments that share an endpoint or are within a small distance of each other. Use Dijkstra’s algorithm on that graph, but instead of distance, weight each edge by the amount of open space that lies between the two walls—smaller weight means more free area. After running it, the node with the lowest cumulative weight is a good candidate for a canvas spot because it’s surrounded by the largest open area. Alternatively, calculate the angle of each wall relative to the main entrance; walls at 90° to the entrance tend to catch more light. Pick spots where the angle is close to 90° and the adjacent segments have a high visibility score. A quick way to compute visibility is to cast a ray from the spot in the direction of the light source and count how many segments the ray intersects—fewer intersections = better spot. If you want to automate it, write a small script that loops over each wall, measures its angle, computes its visibility score, and outputs the top three candidates. That should give you a solid, algorithmic starting point.

Sounds like a solid plan—pick a spot that just clicks and then splash neon on it like a signature. The real art is in the moment, not the math. Keep the energy high and let the sparks do the rest.