jfheath wrote: ↑Mon Sep 05, 2022 6:52 am
But I do know that the maps in North Carolina can cause some weird behaviour - they seem to have roads plotted that disappeared years ago.
[/i]
This is common in the USA. I believe that it has to do with how the map database was created and how it is maintained. This could be a long and complicated discussion/explanation so I will try to keep it simple.
I am going to take a SWAG here and say that the original map database was and is potentially built/maintained by a collection/compilation of thousands of map databases in the US. These most likely have come from the county level in each state. The US to the best of my knowledge does not have a standardized road network database covering every state at all levels of road classifications.
Many if not all of these county level databases were created in the early years of digital mapping and have been maintained over the years using a multitude of methods. Many secondary, urban, country roads etc. may or may not exist in reality. This is dependent on how the data was entered and maintained. Many roads get added from land developer surveys and if the land development project takes many years the road in reality may not exist but does exist in the map database. I know for a fact where I am at in Arizona that there are many roads that are shown even in Google maps that do not exist, but the routing in Google will still try to using them. It is very annoying. Higher level roads Interstate, state, county level roads are well maintained.
I am going to venture a guess and say that there is a priority level set as to the order of what roads get up dated in the map database based on the road type. Who does these updates is still very unclear since there are several sources of data for maps across the county and all are similar but could be different at some level of detail. As an example in basecamp and Google I have found many times that there are different names for the same road. This typically happens at the lower level of the road priority, urban streets, non-county country roads and dirt roads.
I typically use two sources of map data when creating routes in Basecamp for use in my SatNav units. This is where I will find discrepancies in map data and maybe have to use Google Earth for a better view of road surface and such.
From my experience if I cannot see a road track with satellite imagery is on in Google/google earth the road either don't exist or it is very under used and does not show well in the imagery. Typically these would be roads that are dirt/two track.
It is still baffling as to how of this data is maintained, especially at a global level. If one is to look at the number of the road segments in the map databases that are added and maintained on a regular basis it is astounding how good the data is. There are millions of road segments that are in these databases and all of the underlying attributes that are required to do what we want these systems to do. And then there are all of the other data components that are maintained all POI's, address data, directional data, etc. If we are looking at error rates for every million data elements with a .01% error rate that is a 1000 errors in data elements per million elements. How is this level of accuracy maintained on a continuous level. As a qualifier I am just using .01% as a guessed level of accuracy because I don't really know. But still as users we will most likely not see many of these errors as they are not available for our visualization but are used by the routing algorithms.
I don't think as users we understand the complexity of what goes in to what our mapping and navigation systems go thru to make them usable for us as users.
