A Contemporary Nolli Map: Mapping urban public space based on OpenStreetMap Data
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Alternative Title |
| |
| Title of Series | ||
| Number of Parts | 266 | |
| Author | ||
| License | CC Attribution 3.0 Germany: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/66334 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
FOSS4G Prizren Kosovo 2023259 / 266
10
17
23
44
45
46
47
48
49
50
52
53
80
84
85
91
110
116
129
148
164
167
169
173
174
181
182
183
186
187
199
202
204
206
209
215
241
248
265
00:00
HypothesisLevel (video gaming)Open setSpacetimeWordProjective plane
00:49
Level (video gaming)Observational studyHypothesisProjective planeComputer architecture
01:13
MereologyLevel (video gaming)Spacetime
01:51
BlogObservational studyWordLevel (video gaming)CASE <Informatik>AreaShared memory
02:29
Execution unitMultiplication signOpen setCASE <Informatik>AreaField (computer science)Observational studyCartesian coordinate systemMapping
03:07
Different (Kate Ryan album)Performance appraisalEndliche ModelltheorieCartesian coordinate systemSpacetimeDimensional analysisComplex (psychology)
04:04
Multiplication signAreaTerm (mathematics)UsabilityContext awarenessSpacetimeProcess (computing)Computer animation
04:38
FreewareMobile WebVapor barrierSpacetimeTerm (mathematics)
05:11
Installable File SystemMIDICASE <Informatik>Open sourceSpacetimeMathematical analysisProcess (computing)GeometryOrder (biology)Element (mathematics)InformationData analysisReal numberMultiplication signMetreResultantLink (knot theory)Observational studyHierarchyType theorySet (mathematics)Level (video gaming)Descriptive statisticsSlide ruleAreaAttribute grammarAmenable groupMereologyComputer animation
07:41
AreaLevel (video gaming)Element (mathematics)MereologyCASE <Informatik>Observational study
08:01
BuildingVapor barrierAreaHierarchyBuildingLogic gate
08:28
AreaBuildingInformationElement (mathematics)Process (computing)Vapor barrierWind tunnelWord
09:17
BuildingGrass (card game)System identificationSpacetimeGrass (card game)BitAreaLevel (video gaming)Open setType theoryElement (mathematics)
09:59
Type theorySpacetime
10:17
Data typeBuildingGrass (card game)System identificationInformationGeometryProcess (computing)Level (video gaming)
10:44
CASE <Informatik>BuildingSpacetimeAreaDistribution (mathematics)Observational studySpacetimeExecution unitContrast (vision)AreaDifferent (Kate Ryan album)InformationNeighbourhood (graph theory)CASE <Informatik>Graph coloringPairwise comparisonWordBuildingScripting languageObservational studyLevel (video gaming)Type theoryElectronic mailing listBasis <Mathematik>Image resolutionDistribution (mathematics)Mathematical analysisBitComputer animationEngineering drawing
13:28
CASE <Informatik>HypothesisObservational studyCore dumpMappingLevel (video gaming)Multiplication signResultantAreaField (computer science)Scripting languageInformationBitMereologySet (mathematics)Computer animation
Transcript: English(auto-generated)
00:08
I will give you some insights into my master thesis project, which is about mapping urban public space based on open street map data. And before we dive into it, I just want to say a few words about myself.
00:23
As I said, I'm currently working on my master thesis, and I have a background in social sciences. And in the past year or so, I have been focused more on GIS and cartography. But the research I'm doing is still very much motivated by questions of social equity, about how space is distributed, who has access to space, and also what is the importance
00:45
of public space for the urban community. And when I was looking for a master thesis project, I was inspired by the nollie map, or so-called nollie map. It's a historical map from 1748 from Rome by an Italian architect and cartographer Giovanni
01:04
Battista Nolli. And especially in urban studies and cartography and architecture, it is very well known for mapping public and private spaces, and for doing that in very much detail. So here's just a small part. The whole map is huge.
01:20
And you can see the white parts, which are generally interpreted as representing public space or publicly accessible space. And then you can see the dark parts, which are interpreted as private space. And for example, on the left, you can see that Nolli, he even mapped indoors and rooms. In this church, for example, he mapped the room inside as white, but then also the
01:44
walls and the columns in dark, because they are like concrete and not accessible. So I was thinking it would be a good idea to combine, like to bring the basic idea behind it, to bring it into nowadays and to combine it with OpenStreetMap.
02:01
And I'm just assuming all of you know that base data set, I'm using OpenStreetMap at least more or less. And I'm not applying it to Rome, as Nolli was originally doing it, but I'm doing two case studies in Vienna, because that's where I'm currently based. And one of the case studies you can see also on the map is where I'm going to show
02:22
some examples today. So I'm just going to share a few words about this area so you get some kind of idea. So this is the case study area that I was looking at. It's an area around Raabenhof, which can be translated to Raven's Court, which is
02:40
a quite big communal housing unit in Vienna's third district. And most of the area, the case study area that I looked at, is residential area. And the whole size is 500 times 500 meters. And within that area, I did also a lot of field mapping, I added data on OpenStreetMap based on walking there and taking pictures, and also based on some open government data.
03:06
But before we actually get into the application of whatever I was doing, the main question is, what is public space? And it's actually quite a complex question, and it's quite difficult to define.
03:21
On one hand, there's the everyday understanding. I'm quite sure we can all agree on some common understanding of public space, maybe without even being able to actually define it. But then there's also the scientific research about public space, which is a lot about evaluation models and different dimensions and going in a lot of detail.
03:42
And then there's also these general definitions, which can be helpful to get a general idea of public space, but not so helpful maybe to actually identify public spaces and classify a space if it is public or not. So these are some of the challenges that occurred when I was trying to get a definition
04:00
of public space for my application. So what I ended up doing and what I ended up deciding for within my context is to identify or define public spaces if they are publicly accessible and usable areas, and usable in terms of I can go there and I can spend time there, and they actually can be used
04:22
as public spaces. That's also why I'm not considering traffic areas as public space. And I'm also only considering outdoor spaces. And the whole process I was working on, it only really relates to the urban context. So just a quick overview of what I exclude, what I do not look at, and just to get
04:43
that clear already, as I already mentioned, I'm not looking at indoors. I'm excluding anything but ground floor. I also do not look at accessibility in terms of free from barriers. For example, with restricted mobility issues. And I'm also not looking at the actual quality and use of a space.
05:03
It's more about potential public space, what might be used as public space or could be used as public space. So how does it work now? We have OpenStreetMap and when I get to some kind of contemporary Nollymap, and I've developed a process for that, and some Python analysis, which is also available as open
05:27
source on GitHub. The link is down here and also in the talk description, and it will be on the following slides, so no rush. So what it does, it takes an extract of an OpenStreetMap data set. As I mentioned, I use case studies with a size of 500 times 500 meters.
05:45
I load these as shapely geometries with some additional attributes about the tags and ID of OpenStreetMap and these things. I do some data cleaning and filter out some elements. And then actually the real data analysis starts, where I process both the tags and the
06:02
geometries of the OpenStreetMap elements in order to identify two things. The space type and the public access. And then the result I get is a GeoJSON with geometries that have these two informations, the public accessibility and the space type, which can then be
06:20
visualized, for example, in QGIS. And for sure, one of the most interesting parts, I would say, is the processing of the tags and geometries. What did I actually do there? And especially the public access analysis is based on three layers, which are in some
06:43
kind of hierarchy. So, I chose what sources of information I can use from OpenStreetMap for the public access, because it's not just there as a tag usually. And so, I've developed this hierarchy, which has three layers. And the first layer are tags that describe or indicate public access.
07:05
And I count that as the most reliable source of information, because it's very explicit. And I use different tags or I analyze different tags in this step. Of course, the access tag, which can describe private access or public access or access
07:21
just for residents or customers or whatever. And I'm also looking at some other tags that indicate restricted access. So, for example, if an amenity or area, it has restricted opening hours or there's entrance fee related to it, this is interpreted as restricted access.
07:42
And here you can see a part of the case study area and the elements that have one of these tags that I'm looking at on the first level. And it's not too many, as you see. So, for the elements that I wasn't able to identify access based on the first layer
08:00
about the tags, I'm moving to the second layer of my hierarchy, which is inaccessible enclosed areas through barriers. So, I'm looking at barriers like fences, walls, but also rails and buildings. These are my barriers from OpenStreetMap. And then I'm also looking at entrances to these barriers or through these barriers
08:24
like gates that are accessible or footways or crossings. And combining these, I get the inaccessible enclosed areas. So, what you can see here in blue, these are the buildings. Without, what's the word, building tunnels.
08:42
They were cut out, if you know what I mean. And barriers like fences and walls. And then everything that is within these barriers in red hatching, that's the inaccessible enclosed areas. And then I'm using the information of inaccessibility that I have in this red hatched areas.
09:02
And I'm transferring this information to the OpenStreetMap elements that either are within these areas or intersect with it. So, there's some geometric processing happening to get the public access information from here. After these two layers, there's still quite a bit of the area that I want to look at
09:23
that I don't have identified public access yet. So, then the third layer comes to work. And where I'm deriving the access from the space type. So, for example, if there's a land use tag for grass in the OpenStreetMap element,
09:44
I assume grass is generally publicly accessible if it's not within an inaccessible enclosed area and if it doesn't have access no tag or something like that. Because these are the two layers before that. Or, for example, if there's a leisure outdoor seating tag,
10:03
I derive the space type after seating and this is interpreted as restricted public access. Because usually it is publicly accessible, but it is related to having to consume something. And that's already some kind of restriction. Yeah, as I said, for grass, for example, it is assumed that it is publicly accessible.
10:25
So, this is the whole process. And then at the end of this, and some more cleaning and processing, there's this geojson with the geometries and the information. And this can be visualized. And I'd like to show you how I visualized it.
10:40
As I said, inspired by the historical nollie map. So, here you can see, again, the case study Rabenhof, Radens Court, with white areas which have public access, gray areas with restricted public access, black areas, no public access. And then, as I said, I'm not looking at buildings, so these have undefined public access.
11:04
And with the colors that are also more or less taken from similar works about the nollie map and the original nollie map, it's quite a strong contrast between public space and not publicly accessible space. And what's interesting in this case study, for example,
11:22
is that there's quite a bit of publicly accessible space within courtyards, because there's a lot of building passages, I think that was the word I was looking for earlier, in these typical communal housing units in Vienna. So, then the courtyards, they offer public space to the neighborhood, which can be nicely seen here.
11:42
As mentioned, the resulting geojson does not only have the access information, but also the space type. So, the space type can be added as another layer to get a more differentiated picture of the public spaces. And then, for example, you can see lots of inaccessible space, it's actually traffic area,
12:06
but there's also some playgrounds, for example, which are under play and sports, and which are inaccessible, and some are accessible, and some are restricted. So, this can all be, like the resolution is pretty bad, but I hope you can see some of it. And it allows for comparisons between different neighborhoods, technically even cities,
12:27
though the Python script I've developed is not focused on performance, so a whole city will take really long to process, but it allows different comparisons, and also to get information about
12:42
how the space is distributed within an area. And this becomes even more visible when comparing the area distribution as a bar chart, for example, where it becomes more visible, like how much space is actually accessible and how much is not. And you can see, as I already mentioned, lots of non-accessible space, traffic area,
13:04
and in this case study, for example, it's almost one third of the area is accessible, not accessible, and undefined access because of building. And it allows different analysis about this area,
13:20
but as you're probably not familiar with the case study area, this is just to show you what can be possible, and I want to finish with my key learnings. First, which I already mentioned a bit, is that defining public space is pretty complex, especially in a very concrete application, and in the end, I chose public accessibility as the defining criteria,
13:43
which worked out fine, I would say, but you have to keep in mind, it's only one aspect of public space, and there's a lot missing still. And then the second learning is that a contemporary nollie map is indeed possible with OpenStreetMap, which I think is pretty cool, but you also have to keep in mind that I did spend a lot of time
14:02
on field mapping and adding data to OpenStreetMap to get a reliable result about these case study areas, and I would love to hear your opinion about that, because I'm not, like I've just started with my master thesis to become part of this whole community, and I think that the most important tags and information that I'm using,
14:24
like the barriers, fences, and the access tag, it's usually not the core of OpenStreetMap, and that's why the dataset, probably in most cases without adding data, is not sufficient to just run the script and then have the map. So there's some more work involved, unfortunately.
14:43
So that was it, thank you very much.