There were relatively few tools for producing good quality overlays, especially with large datasets. Google maps overlays are great but limited to ~10,000 points which forces you to summarize and split the data.
I made a few apps that pull the dataset from S3, create the heatmap overlays (100,000+ pngs), optimize them, zip into batches and upload them back to S3 for the app to use.
I've looked recently and there still aren't any easy ways of producing overlay images - there is openheatmap but the images it creates are fairly ugly and the user experience appears to be pretty poor.
https://www.dropbox.com/s/uls195x89r57rh7/ss_3.5_all.png?raw=1 is an example of the sort of heatmaps I'm producing but it would be simple to produce more overlay styles.
For v1, I’d make a user friendly way of loading data and demo the overlay on google or apple maps. Producing the images takes a lot of processing so eventually I would autoscale but for the POC I was planning to limit it to a fixed set of servers.
I could offer apps that let you pick your dataset and see your overlay in a real app. I'd also provide code for iOS, Android and web.
I was planning to charge (S3 + bandwidth) * a small markup initially but this could be plan based later.
I'd expose an API so it could be automated.
I'm not sure how to work out if people would pay for this service - it's a fairly niche thing but I would have used it for my app had it existed. I think it could be useful for journalists or anyone plotting large geographical datasets.
I'd estimate I could get a functional and acceptable looking MVP up in around 2 weeks.
My question is this really: is it worth, or how do I work out if it's worth, building this service?
Thanks in advance for any advice or opinions!