Maps

I've been active in the OpenHistoricalMap (OHM) project for quite a while now. We've had our ups and downs, but are busy recovering from a major server outage and things are looking better and better for the future.
My early work has been in mapping historic race tracks in North America. There was some complexity in locating them and finding maps and or imagery to allow for mapping them in OHM but ultimately, you dropped them in place and set the start and end dates for their usage, and it was done.
Things have gotten a lot more complex recently. I've been working on mapping the history of the Capital District of New York, starting with Albany, with plans to radiate outwards and map Troy & Schenectady.. This has made things a little interesting. There is a relatively recent episode of Doctor Who in which all of history happens all at once. If you are going deep in time in an OHM project, the editing process consists of trying to map all of history all at once, and the more detail you map, the more challenging the editor becomes to work with.
There is a filter mechanism in JOSM (the OSM editor I use) that helps to some degree - I have a limited ability to ask for things that started or ended before or after certain years - but it could use some refinement. OHM's approach to dating could use some work as well; the Gregorian calendar stops in 1532, and older things may have some uncertainty about when they actually happened. For example, no one is entirely sure on when the northern end of the Albany Basin was filled. We know it was before 1910 and after 1892 but there's a bit of guesswork involved (and the answer may be in some forgotten records in the Albany city archives.)
The other project I've started on is the Antietam Battlefield from 1862. But that's a topic for another posting.

I have been an active contributor to OpenStreetMap since 2009, and was a member of the founding board of OpenStreetMap US. I have taken two years off from the board, and now feel ready to return.
I am extremely impressed with what the board has accomplished during the time that Martijn has been President, and look forward to supporting those efforts going forward.
I think one of the critical efforts needed is improving the "on ramp" for OpenStreetMap. We've been talking about this for years, but the progress made has been limited. This is still a very intimidating project from an outsider's point of view.
Disclosure: I am currently working as a consultant at Mapzen.

This is all Android based stuff. A previous generation Nexus 7 tablet running OSM Tracklog and a repurposed HTC cell phone running OsmAnd. Click for bigger:

...aren't necessarily useless after the contract is over and you've acquired the new one...

i have recently repurposed an old iPhone 3GS as an iPod touch, something it does pretty well. and my wife's retired HTC Android is now serving as a GPS, thanks to OsmAnd, an OpenSource GPS program for Androids that uses OpenSreetMap data. so don't assume that the old phones don't have a purpose any more, they can still slurp up and export data via wifi, and can run useful programs.

what i have learned, though, is that usually they won't come up unless they have a sim. the good news is that any sim will do, even old ones that have been deactivated by the carrier. put the sim in, "activate" the phone, take the sim out and put it in airplane mode. this works similarly on iphones and on androids in my experience.

Jimapco is kind of a legend in the Capital District of New York. For the longest time, they have been a notable supplier of maps of very good accuracy (not perfect, but very good). They are under copyright, of course, and so are not acceptable sources for OpenStreetMap, but I have always respected the quality and when I used to put on TSD rallies I would always buy a couple of the relevant county maps each year and hand them out to my checkpoint crews so that they'd get to the right places.

So last night I was at Tech Valley High School for a chain of meetings involving various combinations of parents and students, and opted out of the social committee meeting (as that was really my daughter's gig). I found myself looking at a 2008 version of the Jimapco greater capital region map which happened to be posted on the wall, and decided to look at one or two spots where I knew where there might be issues. I was more than a little surprised.

There were three specific areas where I caught things that were significantly out of date. The first has to do with a cluster of town roads at the Sand Lake/Nassau town border on the east side of Burden Lake. If you look at this link (http://www.openstreetmap.org/#map=14/42.5916/-73.5414&layers=N) to OpenStreetMap, you can see CR 20 and CR 47 just kind of end, connecting to town roads. At some time in the past they certainly extended further, but today they just end. The 2008 Jimapco map shows them continuing - but I've lived in the area since 2000 and I'm quite sure that it's been a bit longer than that since these county routes were through. Being an OSMer and a local, of course, I've visited these roads and taken GPS coordinates at the exact spots where county maintenance ends; there is no doubt about this.

The second is the routing of NY 338. Originally NY 338 was a very short highway in Saratoga County, a bypass around Schuylerville. In 1980 it was turned over to the county and was redesignated Saratoga County CR 338. NYS then reused the designation across the river in Washington County for a short state route that passed through the hamlet of Cossayuna on Cossayuna Lake, connecting with NY 29 on one end and NY 40 on the other. The 2008 Jimapco map shows this alignment. The problem that NY turned the road back over to Washington County in 1996 and it's been Washington County Route 49 since then: http://www.openstreetmap.org/#map=13/43.1635/-73.4490&layers=N. So that's an update that's 12 years overdue in 2008. I was putting on Road Rallies in this area back in the '80s and I certainly remember when this was NY 338 - but I've been back since and it's definitely not NY 338 today.

The third is the routing of NY 66 in the vicinity of the hamlet of Averill Park. This link to OpenStreetMap shows the modern routing, which passes east of Averill Park, through the hamlet of Sand Lake (all of this is contained within the Town of Sand Lake): http://www.openstreetmap.org/#map=14/42.6403/-73.5544&layers=N. This particular routing dates from 1980. The 2008 Jimapco map, however, shows the pre-1980 route, where NY 66 passes through the center of Averill Park. The modern CR 45 is Old Route 66, which meets NY 43 in the center of the hamlet. In the old routing, 66 and 43 overlapped from this point east to Sand Lake, where both then turned south (right). So in 2008, this update is 28 years over due.

So this is really a very disappointing performance from a highly regarded map provider. I suppose I should run up to their retail map store and see if the current version of their map still has these problems, and maybe look at some of their more detailed maps (say, the county maps) to see if they share the problems.

i've decided to start using my OSM diary for a lot of the mappy stuff. When I post new articles there, i'll provide links here.
My latest, on the progress of Highway Shields in NY, is Shields Up. a list of all diary entries may be found here

I have gotten a lot of good use out of entry level Garmin Nuvis. What i haven't gotten is long life. Typically i get maybe 2-3 years out of one before either the slider switch on the top fails or the USB connector fails. These failures are generally out of warranty and the only thing Garmin offers is an exchange for a refurb unit that is close to the cost of a new unit.
I don't recommend paying for lifetime maps on Nuvis. The lifetime of a well used Nuvi is simply not that long.
Having said that, i will probably replace the current Nuvi (failed slider switch) with another Garmin when the time comes, but i may consider going a little upmarket next time out.

One of the issues that will come up in these projects is local variation. There will be differences on a state-by-state basis, and unique things within a jurisdiction. The only way we're going to learn about these is by pushing forward, but we can think about them in advance so that there's a plan.
Each time a new department or company is brought into the process, there will be a discovery period. Local statutes and policies will need to be figured out. There will be important items that are not necessarily appropriate as data for OSM. Thus, we will have the local policy file.

I don't know what form this will take yet. I'm on a learning curve with mkgmap. I don't think there's an "emergency" mode for a Nuvi, and there are multiple classes of emergency vehicle. An ambulance is not a police car, is not a pumper, is not a hook-and-ladder truck, is not an extensible ladder truck. Each different class of equipment may have a different set of routing constraints. Some of these constraints can be configured using traditional OSM tagging (maxweight, maxheight, maxlength), but others may reflect specialized local knowledge. We are very likely going to end up with different maps for different equipment classes.
We also must incorporate specific local knowledge. The local responders may be able to tell us that certain routes should not be taken if at all possible. We will need a way to describe that and get it into the GPS of choice.

Now, and this is IMPORTANT: The primary application for this is not for a particular engine company or volunteer district responding in their own territory. The primary application is a mutual assistance call, where, say, the McKownville VFD receives a request for assistance from the neighboring Westmere VFD and needs to go somewhere they ordinarily don't. Whenever I describe the project to someone involved in the real world of firefighting, they invariably bring this issue up without any prompting on my part. Out of area response weighs heavily on the minds of firefighters.

So there's one of the key areas that needs to be fleshed out -- how do we reflect local policy? Right now, I'm interested in ideas and approaches, not code. It's not time for code just yet. Have at it in the comments.

Some recent discussion on the OSM tagging mailing list brought something to my attention that has to be a part of the Emergency Services Mapping project.
The additional component is indemnification. OSM and the cadre of Volunteer Mappers have no resources to defend a lawsuit. There will be a need to provide protection for them in our modern legal climate. I'm adding this to my presentation and will have to insure that protection for them is part of any formal project that gets launched.

This is a project I've been thinking about since I started seriously reviewing the map of Albany, NY in OpenStreetMap this summer. This blog posting is based on my lightning talk at SOTM US 2012, but expands on some things. Also, there has been some change since I gave that talk a week ago and those changes are reflected here.
I was working through the Pine Hills neighborhood and it occurred to me that responding to a fire call in Albany could be pretty challenging. There were some pretty significant issues in the Tiger 2005 data. After finding and correcting issues I discovered on the road, I would in some cases go find the corresponding area in Google Maps, and could see that they'd found and fixed some problems and not others. A good example is Helderberg Avenue. This first link is my corrected map as reflected in OpenStreetMap: Helderberg Avenue in OSM. In Tiger 2005, Helderburg Avenue is shown as one continous street, and my correction to it consists of breaking the street to reflect the two pedestrian only sections. The second link is to Google Maps; as of 10/20/2012 the Google Maps data has partially corrected Tiger 2005 -- it has one of the pedestrian only sections, but not the other: Helderberg Avenue in Google Maps.
My review of the Garmin map of Albany suggest that the Navteq data being used by Garmin is a bit better, but still imperfect. But I've seen Garmin take 2 years to pick up on significant changes to the road grid; they didn't have the new bridge over the railroad yard in Rensselaer for quite some time after it was completed and opened, and that was a major alteration to traffic patterns.
So the gist of my first argument is that commercial map vendors don't really seem to have the precise maps needed by first responders.
The question then becomes this: Can OpenStreetMap provide such maps?

I did a presentation on HOT for Albany Rotary last Wednesday (2012-01-18). I just published the slides, they're available Here.
HOT is a remarkable exercise in applying modern technologies and community based approaches to an unexpected but extremely effective approach to Humanitarian aid. HOT applies the tools and methods of OpenStreetMap to humanitarian efforts, and leverages the OSM community to crowdsource the responses to disasters like the Haiti earthquake. HOT also engages with local communities in the developing world, enabling them to map their communities and take control over their own fate.
HOT is a remarkable effort that (from my point of view as president of a Rotary Club) fits in well with Rotary's approach to international projects. I am planning to continue promoting and advocating HOT in the Rotary Community, starting out in District 7190.

As an active member of the OpenStreetMap community, I view the appearance of Google MapMaker with a mixture of interest and apprehension. Interest, because Google frequently (although not always) does these things very well, and apprehension because I'm deeply invested in a project which without question is a direct competitor.
So what are the issues as I see them? First of all, Google keeps using the word community. This is what I think are the elements of a true online community project:

• Community contributions


• Community governance


• Community ownership

So consider these elements, decide if you agree or disagree, and then look at what Google is doing, which involves only the first of these, and decide if what Google is doing is truly a community project or not. Me, I don't think so.
A secondary issue, as I just popped into MapMaker to fix one known error in West Sand Lake (a missing chunk of Helen Lane, see here: http://g.co/maps/d6yeb), is that I never even saw the terms and conditions, I have edited the Google Map without ever agreeing to anything. This was probably a side effect of being already logged into my Google Account. It seems to me that this isn't much different than a shrink wrap license. Oh well...

i'm highly amused that we're at the point where a new entry level GPS is cheaper than a map update for an old entry level GPS.

Here they are: http://www.flickr.com/photos/nfgusedautoparts/sets/72157625786459692/.
They are public domain, don't let the flickr CC copyright notice fool you. Flicker didn't offer me a public domain option for setting rights.

The issues with Historic Maps at the Library of Congress turns out to have a happy ending.
The Linux tool to convert them from MrSID to jpg is available from LizardTech, and while it didn't work in an old copy of Fedora, or a current copy of CentOS 5, it did work in Ubuntu 10.1. So I have converted all 16 images from the LOC site to jpg.
The other good news is that as a product of the US Government (not to mention having been published before 1922), they are public domain. I'm uploading full sized jpgs to Flickr right now, will post a followup when they're ready.