I need advice about geolocation services and geocoding

So first of all greetings. I´d like to ask some advice about geolocation and your experience with different services available.

Full disclosure: this is a non graded project (to research and learn) for my college coding club. We´re working with a research team of electrical engineers to create a proposal for a system of electric vehicles (green project) and the part we’re working on is planning the setting off charging stations across the city.

My proposal was to create a mobile app to track (with consent) the travel routes of a number of taxi drivers across the city. This will be done by periodically logging the geolocation and the hour/time of the vehicle/mobile with the app and registering in an internal database (SQlite) and dumping the day info daily on Firebase (or something similar) to later process the most common route intersection points (to then set up the charging stations there) …or at least that’s what i figured could be done, now the hard part is that i don’t know how to do any of that (hence research) so any advice on these topics would be welcome:

  • As far as i can tell this is basically a fleet tracking app but since the requirements are that we build everything from scratch (this IS a learning project for college) AND that we keep cost to minimum (free tier preferable), i’m not sure if the free tier of Google Geo API is what i need (honestly i’m a bit confused on their pricing plan criteria), so any advice or your experience with geocoding will be welcome.

  • I’m assuming that the geo call will be done on a 30 second interval over a period of 8-12 hours, this is 960-1440 calls daily per vehicle and if i wanted accurate data i will need at least 20 vehicles (is a small town all things considered), so around 19200-28800 calls daily all told …which wipes out the free tier of Google Geolocation in 2 days flat (if i got my math and understanding of their service correctly) Is this excessive? is there any service that can do at least half this much on a free tier?

  • I think this can be done with an absolutely minimalist android Geo API app, but i haven’t studied mobile development yet so i may be wrong (another thing to study here).

  • The data processing could be done by someone else but i may need to do so myself so any advice on how to do that will also be welcome.

…I’m probably not ready to do this kind of thing, I know my HTML/CSS/JS (basics) but we’re just starting PHP and dbs this semester so while i kind of know what could be done i have no idea on how to implement this project. I humbly ask for advice on how to start, :sweat_smile:

Thank you for your time and have a nice day.

P.D: Originally posted on Codecademy forum, figured i’ll try to hit both communities and see what kind of info i get but if it goes against any policy let me know.

So if you do this for 30 days, you need around 1Mio API calls.

The Google billing page states 0.004 USD per each request, so this would result in 4.000 USD per month.

So it’s up to you if this is “excessive” for your project.

Google seems to be mostly the most expensive solution,
that’s why I would search for:

  • TomTom
  • Mapbox
  • Here
  • Bing

I think in the end you have to pay at least 500 - 1.000USD for these services.
Sometimes you have to invest money in cool stuff.

What level are the research team? Because if it’s PHD research, your faculty can apply for credits up to a value of $5000, see:

As @miku86 says, do check with other providers, as well as Amazon. Particularly, check their price policies w/r/t education/research – even with Google, there’s a specific tier designed for education that give more credits than the basic free tier.

Also, is 30 seconds enough? It seems high for telemetry, are you going to get data that can be accurately mapped to roads with that high a period between heartbeats?

No it’s undergraduate research hence why i’m out of my depth and the 30 secs was more a measure to reduce the number of calls i’ll need to use - and thus the overall price for the geo api. I figured if i account for traffic that latency might be tolerable enough.

Yeah i figured we wouldn’t be able to escape at least a basic budget but I want to at least try to make the prototype on a free tier and then evaluate if it’s cost effective enough to move to a budget.