Crowdsourced Maps Project - mods help/thoughts

**malcom2073** · 10-05-2009, 03:36 PM

The purpose is the production of open source but commercial grade algorithms similar to the ones already in use by some organizations (internally) to parse GPS tracks which are to be automatically collected and submitted with the end goal of generating valid and accurate routeable map data.

If I understood correctly that is.

**Heather** · 10-05-2009, 03:45 PM

I think the purpose is clear - free maps, and free navigation, eventually. Bugbyte tited this fairly well as the FreeNav Project - i think it works, no?

We are facing a chicken or egg situation, we can keep discussing which data we need, or start by collecting the basic NMEA raw data and then look at that to determine what other types of data we should collect.

The expectation is that we can pool the Open Street Map data, we can also use freely available US data, which OSM used to get started, and if there are enough people actively uploading tracks, we will be able to also attract people from outside the forum to help us improve the maps.

The key is that we are setting out to do this in an open and crowd source fashion.
Step number one is proving that that we can do this project, by demonstrating that we are able to get people uploading the tracks.

Also - should we all agree on where to keep open source software? The other day someone suggested http://github.com/ instead of SorceForge. We should choose one and I can create an account. Thoughts?

**Tical84** · 10-05-2009, 03:52 PM

Am I the only Canadian so far? Should I even participate for now? I live in a city of only 90K and only rarely leave to Vancouver.

**Heather** · 10-05-2009, 04:04 PM

Originally Posted by Tical84

Am I the only Canadian so far? Should I even participate for now? I live in a city of only 90K and only rarely leave to Vancouver.

Yes, absolutley. Not only for the project here, but for all other free and open mapping projects.

**justchat_1** · 10-05-2009, 04:20 PM

Originally Posted by malcom2073

The purpose is the production of open source but commercial grade algorithms similar to the ones already in use by some organizations (internally) to parse GPS tracks which are to be automatically collected and submitted with the end goal of generating valid and accurate routeable map data.

If I understood correctly that is.

Which sounds like re-inventing the wheel. Right now Open Street Maps has 1.2 Billion points (thats not GPS coordinates thats actual waypoints) in its database contributed by a community that is now almost 170,000 users. I would say there doing a pretty good job and providing the data in a completely open source and free way.

Navit is a completely open source routing engine designed to use these free maps.

So what (other then it being mp3car branded) would make this project anything other then re-inventing the wheel.

As is...sequential numbering of the uploaded data makes it useless for most purposes-which is the reason I was pushing to have things decided on before the data is used.

**malcom2073** · 10-05-2009, 04:31 PM

The point is to automatically generate the maps which are currently manually generated. Have you ever tried to turn a GPS track into an OSM compatible map? It's a huge hassle. If we can automate this process (like companies have), then we can generate a huge map base, and possibly contribute to OSM and turn it into a real viable solution. We're trying to contribute, not to parallel.

**justchat_1** · 10-05-2009, 04:56 PM

Originally Posted by Heather

I think the purpose is clear - free maps, and free navigation, eventually. Bugbyte tited this fairly well as the FreeNav Project - i think it works, no?

That makes it seem an awful lot like you guys are doing your own maps and your own navigation.

So is this data getting contributed directly to Open Street maps or is the goal to use open street maps data as a starting point and do a new set of maps?

Second question being where exactly is the missing piece of the puzzle coming from?
Waypoints+Tags=Map (clearly missing something)

I really don't think you guys realize how much your biting off here. Your talking about someone who really knows commercial databases to do the initial data processing and storage. Then you need people skilled in vector manipulation, vector optimization/compression. Then you need people skilled with working with GIS and GDF files. Finally and hardest of all you need a really good web developer that can link everything together in an effortless ajax interface. Were talking thousands of hours even for someone who knows exactly what their doing.

**Heather** · 10-05-2009, 04:59 PM

Originally Posted by malcom2073

The point is to automatically generate the maps which are currently manually generated. Have you ever tried to turn a GPS track into an OSM compatible map? It's a huge hassle. If we can automate this process (like companies have), then we can generate a huge map base, and possibly contribute to OSM and turn it into a real viable solution. We're trying to contribute, not to parallel.

Exactly - I have already let the CEO of OSM know about what we are doing and have multiple times reached out for feedback regarding what kind of data can help them. We are going to make it very simple to contribute gps tracks.

In terms of reinventing the wheel, why not try generating free maps and routing engines here on the forums if people want to.

**Bugbyte** · 10-05-2009, 05:15 PM

Originally Posted by justchat_1

That makes it seem an awful lot like you guys are doing your own maps and your own navigation.

So is this data getting contributed directly to Open Street maps or is the goal to use open street maps data as a starting point and do a new set of maps?

Second question being where exactly is the missing piece of the puzzle coming from?
Waypoints+Tags=Map (clearly missing something)

I really don't think you guys realize how much your biting off here. Your talking about someone who really knows commercial databases to do the initial data processing and storage. Then you need people skilled in vector manipulation, vector optimization/compression. Then you need people skilled with working with GIS and GDF files. Finally and hardest of all you need a really good web developer that can link everything together in an effortless ajax interface. Were talking thousands of hours even for someone who knows exactly what their doing.

Well, I'm not sure that leaving mapping and routing algorithms to professional companies because it is hard or time consuming or both is a very good argument for the main issue:

Map data and routing engines remain proprietary. Adding tracks to OpenStreet maps isn't automatic. Passive mapmaking is possible, navigation corporations are already doing it.

This project seeks to make it easier to collect tracks and add them to maps. That's step one.

Step two can process those tracks and add them to OpenStreetmap if we want, or use OpenStreetMap to supplement the tracks. Or both if someone wants to. Point is, they're free to use.

People have spent thousands of hours reinventing the wheel to make things open source, from operating systems to car pc interfaces. A community project to help make nav data free is precisely the type of project that could easily attract outside interests and help push the whole thing forward.

**lambosprit** · 10-05-2009, 06:40 PM

Originally Posted by Heather

We just have to get started - if we can get 2K forum members uploading tracks..then we can get others excited about helping us solve this problem. I am going to start reaching out to universities as well, to get their input.

lambospirit, we had an interesting request:

Would you mind helping start this? Can i suggest that we make it a wiki article so that others can contribute and we can change it as needed over time?

just tell me where and how and I'll have a go.