To review- State of the Map Conference Session Proposal, Please respond by 10AM PST, Friday, Jan 31

[Katie Urey]

Friends,

Tomorrow I will submit a proposal to present GIS Jammer work at the 2014 State of the Map Conference.  The conference is in Washington D.C.,  April 12-13.

By tomorrow, 10:00 AM, Pacific Time, please let me know if I need to make any significant changes to these elements of the form for the session proposal.

Name:  Katie Urey

Organization:  Oregon Walks-Portland State University-GIS Jammers

Email:  ksu...@gmail.com

Session Title:  Leveraging OSM for Pedestrian Experience Modeling and Analysis

Description (200 words maximum) * Please note this may be used in conference program, if accepted.

The experience of pedestrians and wheelchairs users is different from the experience of motor vehicle and bicycle users.  The pedestrian experience depends on the conditions of sidewalks, curb ramps, and other features that are frequently missing from maps.  Analysis for urban planners and individualized services for pedestrians require an accurate walkway network model.

OSM is a great foundation for accurate walkway modeling: it is accurate and detailed, data corrections are quick and easy, and open data facilitates citizen and advocacy group involvement.  We are building a full stack of open source tools which leverage OSM for walkability analysis.  Functionality will include: acquisition and correlation of additional data relevant to the pedestrian experience, walkway network generation, routing, route-based walkability analysis, and presentation of results.  We plan to make these tools available as a Python QGIS plugin.

Potential uses include impact analysis for improvements to infrastructure and placement of parks, grocery stores, and schools, the elucidation of improvement proposals via interactive websites allowing citizens to examine the impacts for themselves, and real-time individualized web services.

The GIS Jammers group is supported by Oregon Walks and Portland State University.  Our goal is to better understand where we walk and to empower activists.

=============

best regards,

--
Katie Urey
Portland, Oregon
503-228-3083 (land line with saxophone greeting)
503-780-9034 (cell, though not always "on")
ksu...@gmail.com

[Travis Driessen]

This looks really good, Katie! I like how this version includes specific mention of the open source tools, interactive/individualized web services.

I made some slight revisions and additions if you like:   1) took out the use of the word  "users" as applied to bicyclists; which in neo-geography/VGI/2.0 world is a hard concept to pin down and quite contentious, 2) disaggregated "motor vehicle users" term to highlight the two, implied, modes,  and 3) added data "updates" which is an important factor/advantage that OSM has in comparison to centralized (city) databases

-------

The experience of pedestrians and wheelchairs users is different from the experience of motorists, transit riders, and bicyclists.  The pedestrian experience depends on the conditions of sidewalks, curb ramps, and other features that are frequently missing from maps. Analysis for urban planners and individualized services for pedestrians require an accurate and detailed walkway network model.

OSM is an optimal foundation for walkway modeling: it is accurate and detailed, data updates and corrections are quick and easy, and open data facilitates citizen and advocacy group involvement.  We’re building a full stack of open source tools which leverage OSM for walkability analysis.  Functionality will include: acquisition and correlation of additional data relevant to the pedestrian experience, walkway network generation, routing, route-based walkability analysis, and results presentation.  We plan to make these tools available as a Python QGIS plugin.

Potential uses include impact analysis for improvements to infrastructure and placement of parks, grocery stores, and schools, the elucidation of improvement proposals via interactive websites allowing citizens to examine the impacts for themselves, and real-time individualized web services.

The GIS Jammers group is supported by Oregon Walks and Portland State University.  Our goal is to better understand where we walk and to empower activists.

--
You received this message because you are subscribed to the Google Groups "Oregon Walks GIS Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to OregonWalksGISJa...@googlegroups.com.
To post to this group, send email to OregonWalk...@googlegroups.com.
Visit this group at http://groups.google.com/group/OregonWalksGISJammers.
For more options, visit https://groups.google.com/groups/opt_out.

--

Travis Driessen

Graduate Certificate in Geographic Information Systems 2014

Portland State University

travis....@pdx.edu

[Cone, Paul]

How are data updates an advantage that OSM has in comparison to centralized city databases?

 

Paul

---

From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Travis Driessen
Sent: Thursday, January 30, 2014 3:32 PM
To: OregonWalk...@googlegroups.com
Subject: Re: Oregon Walks GIS Jam:: To review- State of the Map Conference Session Proposal, Please respond by 10AM PST, Friday, Jan 31

[Katie Urey]

Hi Paul,

Good question, and one that will be added to the anticipated questions list.

Paul asked:  "How are data updates an advantage that OSM has in comparison to centralized city databases?"

Data upgrades in the OSM are flexible in that corrections can be made by any OSM-editor who is logged in.  Hopefully, only changes made by trusted OSM-editors will persist.

We would like the tools to work anywhere, in any city, and across jurisdiction boundaries.  In Oregon, where Metro, cities and states cooperate, it might be argued that OSM data is, in some places, less accurate. 

However, in some cities, it may be that OSM is clearly ahead, and in that case OSM data would be the only data, at least until a government agency provides data.

In terms of advocacy, it may be that advocates play a role by adding attributes to OSM ways before these attributes work there way into the more tightly controlled codes and documents of a government agency.  For instance, an advocacy might want to do real-time modifications to show where streets crossings are closed by construction, where trails meet sidewalks, where urban paths meet streets.  

I see official-government-agency data as the Oxford Dictionary of Known and Accepted streets and OSM as the living record of an emerging language of pedestrian mobility terms.  Government data record will catch up when budgets and policy also catch-up.  As an activist, I want to let my city know that data is missing, but I do not have the patience to wait for city/state/regional data to be updated.

Again, thank you for the question.  If the session is accepted, the question will be asked.

best,

Katie

[Cone, Paul]

Thank you for the clarifcation Katie.  I just want to be clear that in some cases, OSM data is not as current or accurate as government-maintained data.  Even in Portland.  I have found mistakes in OSM myself.

 

Paul

 

(and for those of you who don't know -- I maitain the street centerline for the City of Portland)

---------------------------------------------------------
Paul Cone
Mapping and GIS
City of Portland, Bureau of Transportation
paul...@portlandoregon.gov
(503) 823-4071

---

From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Katie Urey
Sent: Thursday, January 30, 2014 3:50 PM

[Katie Urey]

Paul,

If you can think of even one specific case, send me the intersection, bounding box, it would help me prepare.  I do suspect that some areas in the Southwest- north of Beaverton Hillsdale Hwy are out of date in OSM.

Katie

[Michael Arnold]

With Travis' changes we are, by my count, at exactly 200 words again :)
Looks very good to me!

(Though I still prefer 'great foundation' to 'optimal' foundation.)

I say ship it! :)

[Cone, Paul]

One specific case I have in mind is where someone was making changes in an area to add a couple of comments to a couple of restaurants, and didn't notice that they also modified a way so it was completely off from where it should have been.  Nobody else noticed it until I discovered it 8 months later and then fixed it.  Government data often has multiple eyes looking at it all the time (think 911 dispatchers, firefighters) and so any error like that would not persist long.

 

---

From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Katie Urey

Sent: Thursday, January 30, 2014 3:56 PM

[Jon Kemp]

As a suggestion don’t forget to consider sidewalks exposed to sun versus sidewalks in the shade.  Many people walking will choose the former.

 

Jonathan Kemp, Principal

---------------

EnCo Environmental Corporation

P.O. Box 1212

Puyallup, WA 98371

Work: 253-841-9710

Email: jk...@encoec.com

 

www.encoec.com

P Think Green! Please do not print this e-mail unless it is completely necessary.

[Scott Parker]

This is a great discussion.  It highlights the importance of our investigations into ways of merging centerline and OSM data.  If we are able to do this it will give us a good base for generating the walkway network model and it will help the city to maintain its street centerline data no matter what format it is in.

 

I think it would be great if Paul can evaluate versions of Michael's program as they become available.  He'll know the most difficult places to merge.

 

Scott

-----Original Message-----
From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Cone, Paul
Sent: Thursday, January 30, 2014 4:18 PM
To: 'OregonWalk...@googlegroups.com'

[Don Baack]

Our here in SW where only 15%  of our streets have sidewalks! we are pushing for extended shoulders, frequently combined with climbing bike lanes.  It is a matter of cost benefit, FHwA studies have show extended shoulders provide 80 % of the safety of full sidewalks for under 20% of the cost.  

My question to you all is are you taking extended shoulders into account in your work? 

Don Baack

6495 SW Burlingame Pl

Portland, OR  97239

Ba...@q.com

503-246-2088. Call if you need to contact me w/in 24hrs

[Travis Driessen]

Perhaps "optimal" is too strong of a word. And a more nebulous concept like a "great" foundation is more appropriate to describe the contribution of OSM data to walkway network analysis. And certainly there are many advantages of city GIS data sets; produced and managed by highly trained professionals.

We have had some really good input on a wide range of discussion topics concerning OSM and City/government produced data in relation to walkway network analysis and to use in urban planning, asset management, and other public service topics. Let's keep it going!

To derive more analytical value, we should attempt break the false dichotomy of OSM vs. City Database data; which I may have accidentally insinuated. There are strengths and limitations in both. And these strengths and weaknesses vary according to applying different criteria such as a) the purpose the data is being used for and, b) the various qualities of the data itself; such as accuracy, validity, temporal, etc., c) etc.

And it is interesting to think about how OSM and City data sources can best complement each other; and why; both in the interest of walkway network analysis and also in the interest to potentially benefit city planners and agents of various municipal service providers.

Some relevant topics:

1) since we are trying to merge city Streetcenterline data with OSM data, there must be vital content or qualities that Street Centerline data has that OSM data perhaps lacks.

What precise information or qualities do Street Centerline data have that  OSM/Tiger data do not?  And how is this useful for our walkway network analysis?

One guess, perhaps the Street segment type (e.g. Arterial, Collector, Local, streets, garages, ramps, and freeways, ) of city data that is used for the generation of walkway network and classification of segment/node types. I am not sure OSM/Tigerdata provides this distinction. On the other hand, Scott and Michael mention that they have generated the walkway network from OSM data; so it seems that it does categorize segment and/or node.


As a converse, since we are also interested OSM data what data or qualities of OSM data would be beneficial to our network analysis?

OSM data has trails and path data that street centerline data does not always have, anyone care to elaborate? Is that it or is there anything else specifically beneficial for walkway network analysis?

And folks from the city transportation world- what OSM attributes and information could potentially be beneficial to complement city database?  For example, does OSM have information about the condition of assets that could be beneficial to your work? Or anything else? 

2) The strengths and limitations of being able to update information in flexible or rigid manners. The flexibility of data updates of OSM is a structural advantage that open databases have over closed databases; that is if you are interested in generating many diverse data types from many diverse data sources, but that does not mean in practice this always generates more up to date information or the quality of the data is to be desired.

And to be analytically accurate; City databases fluctuate between centralized and decentralized in this sense. That is, various agencies (and even contracted actors, etc.) contribute different information, and so functions in a more decentralized sense. It is that the verification process is more rigid, both slowing down the ability update data and in terms of the qualifications of trained professionals who verify the data in some way; which in theory limits data inaccuracies but this is not full proof.

How do city agencies attempt to cross-check the coherency, cohesion, and accuracy of the data sets generated and contributed from segmented agencies and sources; if at all?

And what specific the roles that editors in OSM play? How does the volunteered information get verified? And how do editors become "trusted" (e.g. in terms of the quality of the data that they oversee)?

3) How would merging city street centerline data with OSM data assist the city in maintaining the streetcertline data?

And if this could be done, would the city of Portland or metro affiliates utilize this data in some way?

And where are some of the difficult areas to merge these types of data?

If you got to the bottom of this email perhaps I'm wrong in thinking that its way too long! :)  And perhaps you're thinking we left out the discussion on applying OSM data for walkway analysis in other cities :) 

[Scott Parker]

For the last decade or so the street centerline has been the easiest dataset for researchers to use.  There are certainly other people and groups who, like the Jammers, have used "LOCALID" to tie their data to the street network.  We may want some of this data and others who have used the street centerline may want our walkway network.  That the city has maintained this data in the same way with the same structure for so long is good and bad.  It's good for the reasons mentioned above.  It is bad because it is on old data structures that don't allow an easy transition to routable networks.  Bridges are a particular problem.  Since OSM is a lot younger it does not have this problem.

 

OSM tends to be more geographically accurate at the pedestrian scale.  I suspect much of the street centerline was hand digitized from maps.  OSM data is often created from GPS receivers.  If a street is realigned or an intersection moved the city may not update the centerline file if the geography is not changed significantly and the topology is not changed.  This is fine for many purposes but it can have a significant effect on the automatic placement of walkway corners.  In the very few samples we have, generated walkways based on OSM appear to be more accurate than generated walkways based on street centerline.

 

When the city or academic researchers build a database it usually reflects a consistent measurement practice.  Things like traffic levels and actual speeds are usually measured by traffic engineering standards.  Where this data is available it will almost certainly be attached to the street centerline used by the city.  OSM may have tags for something like traffic level but the data may be wildly inconsistent.  Something like speed limit is likely to be complete and accurate in OSM since little judgment or standardization is involved.

 

The walkway generator can use any connected network of polylines.  We could model walkways along rail lines or along rivers if we want.  Users can use pure OSM if they want or pure street centerline if they want.  If Michael is successful in merging the two we will certainly want to use that.

[Scott Parker]

I thought I had my e-mail bug fixed but apparently not.

 

For the last decade or so the street centerline has been the easiest dataset for researchers to use.  There are certainly other people and groups who, like the Jammers, have used "LOCALID" to tie their data to the street network.  We may want some of this data and others who have used the street centerline may want our walkway network.  That the city has maintained this data in the same way with the same structure for so long is good and bad.  It's good for the reasons mentioned above.  It is bad because it is on old data structures that don't allow an easy transition to routable networks.  Bridges are a particular problem.  Since OSM is a lot younger it does not have this problem.

 

OSM tends to be more geographically accurate at the pedestrian scale.  I suspect much of the street centerline was hand digitized from maps many years ago.  OSM data is often created from GPS receivers.  If a street is realigned or an intersection moved the city may not update the centerline file if the geography is not changed significantly and the topology is not changed.  This is fine for many purposes but it can have a significant effect on the automatic placement of walkway corners.  In the very few samples we have, generated walkways based on OSM appear to be more accurate than generated walkways based on street centerline.

 

When the city or academic researchers build a database it usually reflects a consistent measurement practice.  Things like traffic levels and actual speeds are usually measured by traffic engineering standards.  Where this data is available it will almost certainly be attached to the street centerline used by the city.  OSM may have tags for something like traffic level but the data may be wildly inconsistent.  Something like speed limit is likely to be complete and accurate in OSM since little judgment or measurement standards are involved.

 

The walkway generator can use any connected network of polylines.  We could generate walkways along rail lines or along rivers if we want.  Users can use pure OSM if they want or pure street centerline if they want.  If Michael is successful in merging the two we will certainly want to use that.  I think we can just bypass the issue and let our users decide what data source suits their purpose.

 

Scott

-----Original Message-----
From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Travis Driessen

[Scott Parker]

Some interesting features of OSM data:

- The highlighted street and turn-around are designated as "living street".  Note the little triangle where the lollypop joins the stick.  Stick needs sidewalks but the pop doesn't.  Similar cul-de-sac on the other side of the freeway has a different treatment.

- Walkways around the borders of plazas will generate dozens of short and pointless routing instructions.

- Sparse manually entered sidewalks and crosswalks conflict with generated walkways.

 

Scott

-----Original Message-----
From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Travis Driessen

[Scott Parker]

more ...

 

Note the short walkway segments (green).  One terminates in a street (black line) and the other terminates in a sidewalk (green).  It's tough for the walkway generator to figure out something that will work for both situations.

 

It looks like an error where a driveway connects where there appears to be a barrier (pink line right of center).

 

Scott

[Cone, Paul]

To the best of my knowledge the only work that has been to use the street centerline to figure out ped routing is the work that Scott and the jammers have done.

 

One thing that OSM has that the street centerline does not is turn restrictions.

 

In the past there were grand ideas of tying other data to the street centerline, e.g. turn restrictions, traffic calming, bike streets, but the idea of doing that with at least the latter of those two has never been implemented.  Adding all sort of other attributes sounds good in theory but then you either a) have one large database, i.e. the OSM model, where you let a whole bunch of people edit, because it's too much editing for one person, and then you have to figure out to maintain data integrity, e.g. keep bike data people from messing with the main street centerline data without mucking things up for 911 data exports or b) let everyone make a copy of the data and then years later figure out how to reconcile it all when you need to.  The City has taken the latter approach.

 

Street centerline is run through some automated QC passes, and then exports go to BOEC and Metro, who also run QC passes, and then errors are funneled back to the editor (me).

 

Paul

---

From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Travis Driessen

Sent: Sunday, February 02, 2014 10:58 PM

[Cone, Paul]

If a street is realigned or an intersection is moved then the street centerline data is usually updated because usually we are the ones who approve and often do the work, so we have the plans.  In the past centerline was more aligned with aerials were not as good quality and so often the centerline was more aligned with tax parcels.  But we have pretty good aerials now, and the leaf off ones are especially useful, so we are constantly adjusting data to align better with reality.  Also, the flight plan for 2014 includes 3" aerials, and there will also be another leaf off flight soon.

 

Paul

---

From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Scott Parker
Sent: Sunday, February 02, 2014 11:59 PM

[Cone, Paul]

There is a sidewalk on the south side of the lollipop, that connects to the walkway that goes up to the transit stop.

---

From: OregonWalk...@googlegroups.com [mailto:OregonWalk...@googlegroups.com] On Behalf Of Scott Parker
Sent: Tuesday, February 04, 2014 10:58 AM

[Scott Parker]

See attached.  It would be nice to have this little car shortcut in the street centerline.  There are a couple crosswalks that should be in the walkway network model but there is no street for them to cross.