Permission to use usai and usaus graph data

[cjaniake]

Hello!

Please, I work with logistics, and we have some use cases for the usai (United States Interstate Highways) and usaus (United States Numbered Highways) graphs.

They would be useful for us to be able to build offline systems that can indicate the incidence of tolls on a lane, find alternative routes, calculate distances, etc.

We would like to run some experiments using these two graphs, and possibly to contribute with the project using tracking data.

Is it possible? Thank you!

[Jim]

Welcome to the forum.

For everyone's benefit of a little context, cjaniake emailed me with a query about possibly using some TM data. As it's a busy time in my semester, I asked him to post here so our highway data contributors can ask questions and find out more.

[Duke87]

I personally don't have an issue with anything being used since hey, free open source project, all the data being there for anyone to take and do as they will with is part of the point as far as I'm concerned.

That said, others are entitled to their say over their stake in the IP, so please do not take this as an authoritative green light. Only as one vote in favor of it.

[si404]

possibly to contribute with the project using tracking data.

Is it possible?

Not easily. You'd use the tracking data to improve OSM and then tell us that OSM has been updated with an improved routing. However OSM is mostly right, and tolerances here are fairly high on updating precise point locations because otherwise we'd constantly be chasing the 'OSM drift' for little gain.


I'm another vote for yes on the graphs - it's an open source project.

[yakra]

No problem with me. Open source, etc.

Not knowing the specifics of your use cases, a few things to keep in mind may or may not end up being problems...

They would be useful for us to be able to build offline systems that can indicate the incidence of tolls on a lane, find alternative routes, calculate distances, etc.

Tolls:
Travel Mapping data doesn't contain any info about tolls, so that info would have to come from another source.

Distances:
Our data is pretty low resolution. Mileage is in the right ballpark, but not super-precise.

Alternative routes:
Junctions in TM's data are represented by points on different routes being located at the same lat/long coordinates. This isn't always the case when it should be. Our software flags and logs some of these "near-miss points" (some points will be too far apart to get flagged) in the hope they'll get fixed, but not everything does, resulting in some connections missing from the graphs.

Some routes are one-way only; this some another thing our data doesn't keep track of.

We would like to run some experiments using these two graphs, and possibly to contribute with the project using tracking data.

If the goal is to just run some experiments as you said, maybe to use a small dataset to make sure things are working before scaling up, it may be that these items are non-issues and TM's data could fit the bill. I don't know about your use cases of course.