The API works great so far and the work you’ve done to tie it all together is awesome, however the output seems to be highly optimistic. It seems to calculate routes that are virtually impossible to actually accomplish in the real world. The API will sometimes come up with routes that are only 25 minutes to drive, while Google specifies it as 48 minutes. Real life was 46 minutes.
As far as I can tell the duration/distance comes from OpenStreetMap, but the actual plots of the routes are done via Google Maps API. In other words: the routes generated don’t match the plots on the map. Furthermore they don’t really work in real life because the estimates are way too optimistic. Please correct me if I’m wrong.
We’re hesitant to integrate the API in our systems because we’re not sure if the route generated is actually the fastest and even if it would be, it would probably be far too optimistic regarding time/distance which results in all sorts of planning issues down the line.
Is there any way we can improve the results to get it more inline with the more accurate results that Google Maps produces?