Kerb your enthusiasm
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 |
| |
| Title of Series | ||
| Number of Parts | 70 | |
| Author | ||
| License | CC Attribution 3.0 Unported: 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 | 10.5446/58510 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
State of the Map US 201930 / 70
5
7
14
20
26
28
30
31
33
36
38
40
43
44
51
53
55
56
60
61
67
68
69
00:00
Computer animation
02:21
Diagram
Transcript: English(auto-generated)
00:00
All right, so thanks to Danny and Chris for talking us through hopefully convincing everyone why you should care about the curb I'm gonna build off of that and talk about a proposed solution for how we can start to map curb regulations As important as every other part of the curb is I'm not gonna focus on that at all All I care about is the street edge of the curb and what you can and cannot do there
00:22
So my name is Emily Eris. I am at shared streets and a little bit of background We are a non-profit and we build open source tools and digital governance models to help cities move to help keep cities moving So technology has changed mobility has changed and we want to make sure that cities have the data and the tools and the information
00:41
That they need to plan and manage their streets So I have a slightly different perspective on this because I work closely with various cities that are trying to manage their streets and We also have relationships with different companies and outside of OSM the curb is an incredibly hot topic Where everybody wants a piece of this?
01:01
So what we've been doing is creating a data standard for the curb a way that we can communicate this Regardless of base map regardless of who you are And so as we've gone about doing that we've had to think about how can this be compatible with OSM? So that there's an open way that anybody can participate you don't have to be a city or a company And so we've thought closely about how we could make this data fit and what we've arrived at is that
01:24
We don't believe that OSM is an appropriate place to store regulatory data But we do think that it could hold the building blocks the physical assets that are on the street and we could derive from that Elsewhere, so I'll talk through a little bit about that process and how we came to that understanding And how this works and how you can participate today
01:44
so if you weren't in Chris and Danny's talk just to summarize briefly Mobility is changing the way that we shop and the way that we move around cities has changed dramatically in the last five ten years And so cities everywhere are trying to cope with that They're trying to reallocate right-of-way that used to just be dedicated to parking and not really thought about now
02:05
We're thinking well, how do we allocate that in a way that reflects demands or it today? And so in order to do that cities are starting to think about well What could this look like companies are trying to get a bigger kind of piece of this? Other companies are interested in how they can charge for this
02:21
But nobody can do any of this unless we have an understanding of how this public right-of-way is allocated today So where are we starting from and it turns out that this is actually a really difficult problem Because when we talk about curb regulations, there's not a one-to-one relationship with physical space So your average Street has you know up to four regulations on each side each with a start and an ending place
02:44
And what we're talking about is mapping a legal regulation So we're mapping where a temporal restriction or a restriction about who can do what where at what time? How does this apply on the edge of the street? So the regulations themselves are complicated enough, but the location is actually the really hard part
03:04
So what's going on in the background? Is that a ton of cities across the country are struggling with this problem and they're each doing it in their own way There's been no standard way to collect this data or store it And so we work with some cities like DC that are using computer visioning We work with other cities like Eugene, Oregon that have an intern Kyle who goes out and walks the streets with a notebook
03:25
And it gets the job done But it means that we're all doing it in silos and we're each getting a piece of this that doesn't work for other cities In the absence of collaboration a ton of vendors and companies have sprung up some of which are just doing the surveying But some have a view towards managing the public right-of-way
03:42
And when you have a private company offering a technical mapping solution that really is a backdoor to managing a city's physical space That's a dangerous thing So what we've been working on is an open standard curb LR or curbler friends with Hamburglar for short That basically just creates a way to say
04:00
Here's a common here the way we can solve the technical problem if we can give cities and companies a way to Start to exchange this data very similar to GTFS for transit. Then we're all on an even playing field Vendors can do the surveying. They can give the data to cities in a way they can use And cities have a standard to aim for whether they want to do this data collection themselves or not
04:24
So what curbler looks like is it's a JSON file lives outside of GIS and it's a way that we can take information about something like a motorcycle meter turn it into information that has a geometry so we could display it on a map a location that is base map agnostic
04:41
And then structured information about what is it that's being allowed Who can do this at what time under what payment conditions and priority so that we can resolve conflicts over overlapping restrictions? This creates a machine readable data standard that can be ingested into maps visualization routing engines, etc
05:00
So the goal is if we can create this technical solution in the middle and walk people through how to create it Then companies and cities can start to use it. So that's what we've been working on Why am I here as we've been thinking about this? The map matching that we do is based on OSM We believe that data should be open wherever possible And so we really wanted to say well if we're gonna do this work
05:23
Can OSM already do this? And if so, what can we learn and if not, why not? And how can we make this as complementary as possible? and so Here's what this looks like. Technically you could map curb regulations and OSM right now, but I'm gonna tell you why it doesn't work So right now you could go into the way
05:42
So the road center line and you would tag something as parking lane and add parking conditional you can then start to add kind of all these usage rules on top and there's a hierarchy level so that if you have A no stopping and a resident parking you get parking condition right side of the road level one parking condition
06:00
Right side of the road level two, etc. So technically this exists. It's very very rare There's a link at the bottom of this to a blog post that we wrote that kind of looks into this in more detail and pulls Up stuff from tag info But really the only time I see this is to mark free or permit or paid parking There's some cases less than a thousand around the world where somebody did really go into a lot of detail
06:25
This is actually not Germany and this is Toulouse in France and it is one of less than a thousand Parking regulations that have really been mapped up to three levels Which is not uncommon like my street has three But anyway, so you can see how unwieldy this is and I'm not concerned about the number of tags here that it's always gonna be
06:43
Unwieldy map editing tools can help with that The problem is that it can contradict itself in so many places You can have parking rules on both sides of the road as opposed to parking conditions on the right side of the road There's no information about how to resolve that And so part of this is a tagging issue, but honestly the bigger issue is is how we represent this in space
07:05
So here's an example straight with those four restrictions on it This single Street would then be subdivided into 13 different sections some of which have three levels of rules some have four If you look at that green line, sometimes it's level three. Sometimes it's level four
07:21
This is impossible to map. You can't map one thing without mapping 13 others in detail Realigning geographies suddenly has unintended consequences. I realign the road. I have moved a loading zone That's a problem. You could say well, yeah, it's complicated map it properly
07:40
But this affects every urban road around the world I don't think it's appropriate to say map it properly like this just it breaks down You could say map it on a different geometry But sidewalks curb edges same problem if you start to think about an area You've actually magnified that in another dimension so that doesn't work you could start adding new features
08:02
If that looks appealing to you, I'm sorry, but that doesn't work. So fundamentally, it's not Sometimes maps are fantastic But what we came to the realization of is that sometimes the map is the problem not everything can be represented with geographic coordinates so we draw a line we distinguish between a physical asset and what we call a
08:23
Regulatory geometry a concept that lives on top of physical space like a road closure like a speed limit like a curb usage restriction That don't fit neatly into OSM or any other type of GIS There's some ways that we've been able to shoehorn these into OSM something like a speed limit But it only works because it's not that complicated. You don't have 14 happening on the same road
08:45
And so we really think that we need a different way to represent locations where rules apply So our our approach is to map physical assets that you can actually see and observe and ground truth Put those into OSM, but leave the derivation elsewhere
09:02
So here's what that looks like So a city today could go out and map its physical assets So parking meters parking signs the beginning and end of a curb paint zone. You can see that it's what communicates a legal rule to a To a person on the street they exist they're always there and they do have a lot to do in the longitude
09:21
You can then attach structured data to that point about what rules are in place So hypothetically we could call this curbside marker or any other name In our curve our standard we've come up with a pretty detailed and what we think is a very workable Kind of tagging schema you could apply that to OSM very easily or adapt it if parts of it are not desirable
09:43
Similar to how traffic sign gets mapped, but that's only for movement restrictions. And this is really as far as the data goes in OSM What we think is that points can be processed into these regulatory geometries and updated and kept in a different format
10:01
There's a variety of ways that you could do this mapping today, so I mentioned that Washington DC uses imagery They then go in and edit and they get a geometry out of that Eugene, Oregon uses a Kyle with his pen and paper. We actually went out and we tried a bunch of open source methods We used open data kit slash kobo same difference as well as open map kit and we found that there was location issues with both and
10:25
Kind of interface things that didn't really work Our favorite was field papers and a camera. So we went out and surveyed a downtown We've made this available again on our blog It worked really well and we're starting to work with Bellevue, Washington on using this method But there's a ton of other ways that you could use the measuring wheel
10:41
There's it's it's not inherently a difficult problem if you can capture a photo as well as a location So there's a variety of ways that cities can do this now OSM mappers can do this now But regardless once you've got those points, they've got to be converted into street segments So we've created some tools that are available again on our github That will take those points and turn them into segments either by capturing the
11:03
Information about the beginning and the end of where that applies or if you don't have that by buffering something So a parking meter is 3 meter a parking meter is 3 meters wide and we can kind of turn that into a segment The interesting part is then we've got to reference the street somehow in a base map agnostic way
11:20
So if you were at this conference last year Kevin Webb talked a lot about the shared streets referencing system and how we're working to develop map conflation tools so that we can port data Between maps regardless of what of how you physically represent that Street So what happens is we have a linear referencing system that looks at the location of the start and the end of every
11:40
Intersection as well as the bearing for the first 20 meters as long as those are similar enough A street will be assigned the same shared streets reference ID So we work with this so that we can help cities or help companies to match data to maps Whether they're using OSM or their own authoritative data or a different source So this is a way that we can say we are all talking about the same segment of Street use whatever map you want
12:06
So when we snap things to the street and we turn them into street segments Yes, we represent them on a map using open street map, but more importantly We've created kind of an identifier the ABCD 1 2 3 and we're saying this is how we're gonna refer to Main Street at
12:20
45 to 100 meters and then we attach a structured regulation So the benefits here yes, we're doing an outside of OSM, but it is street linked is base map agnostic We're using open source tools. They are available and at any point of a sign gets moved. You can regenerate this So this is the process and what it ends up with is that we take a bunch of points for a city
12:40
We process them and we end up with this curve LR feed. So again three components We have a geometry up at the top and the geo JSON so we can drop it onto a map We have that shared streets reference ID location And some other data and then we have the structured rule
13:00
So once we've got that this is open data from the city of Calgary that a guy named Sadiq Mahidin has turned into With an API and a rules map for the city. We've taken that and we've adapted it and built off of it and this is 35,000 parking meters and signs for the city of LA And so we're starting to develop tools so that as cities go about either converting data that they already have or creating new data
13:25
They can visualize it and see the value of standardized data So that's what we're working on Kind of overall my takeaways. I think that I was hoping to convince you is that not everything fits neatly into a map We can't necessarily fit these invisible rules
13:40
We can fit the building blocks and there are ways that open street map can participate and be a part of this So that whether you work at a city whether your company or whether you just want to advocate for something in your own city You could start to map this and you can do this today I think this is really important because I think
14:01
Regardless of who you are there should be a way that we could create this data There should be a way that everybody can have a role Cool. And so with that, I think we have five minutes left and I am happy to take any questions Alyssa
15:02
Yeah, so the question was how is this data being used to make our lives better? So for the curb it's really early days like very few cities have managed to map this and it's kind of we're just starting to Push out the standard and now work with cities to create the data But what I can say is that we also have a pilot where we work with ride share companies to release
15:21
Aggregated pickup and drop-off data. So there Washington DC has used this information about where there's really high demand to have access to the edge of the street to create dedicated pickup and drop-off zones so that hopefully you have a better experience as a passenger But also you're not in double parking in a bike lane exactly
15:41
Yeah So it's up to the city how they want to create this But a lot of the cities that we work with are pretty progressive about open data portals Anyway, the data that we showed for LA we got from their open data portal So presumably if this is already open and we have something that's derived
16:04
That would then be open as well What I didn't talk about is that if you're creating data that lives in a JSON, you know You also have a discoverability issue We don't necessarily have a solution to that open street map is open you could you could derive this But where does it live like GTFS you kind of have to think about where you can put it?
16:24
Mark you had a question, but your hand came down
16:53
So I think what we've said is that that physical assets? By all means that can live in OSM and we can demonstrate and then anybody could process it
17:00
I think what we end up with is these curbler feeds that are JSONs And so it's not necessarily an issue of an extra layer because again We've kind of taken it out of the map and put it into something that you can drop into your own map or you can ingest through a rules engine or an API But we have a bunch of JSONs. So I think that's the issue is not necessarily How do we make another version of OSM of curb curb SM?
17:23
But it's more like how do we it's more like how do we have a transit land honestly? And who would host that who would maintain it? Is that a university thing? Is that a is that shared streets? Is that somebody else?
18:03
Yeah, we have actually Indranil from shared streets is gave a talk in parallel on Road closures because that's another one where we need to communicate this cities want to do it in authoritative way Doesn't really fit in OSM. How do we describe that? So we've thought about road closures
18:20
We thought about the curb and then I think it just keeps going from there Alyssa So one of our partners is Ford mobility and they have a UK based team
18:43
That's really interested in you know, how do we map London? That's a pretty difficult area to map And so but what we found is that the referencing system like it it works in that context So if you can start to map match maps then you can generate these IDs regardless and it's really a question of
19:01
Has the way that we've come up with to describe the curb does that make sense in their context? and I think because we were working with European partners and UK cities and I live in a small town because we had that context and relationships with various companies We were able to get a pretty good sense. I think of the different needs that we saw We released a draft form of Kerbler a few months ago kind of as an invitation to
19:24
Collaborate and for people to say this works you miss this. What about this? And we did a bunch of revisions after that based on people's input Like any standard it evolves over time But I think for kind of the five to ten cities worth of data that I've looked through I haven't found anything that doesn't fit
19:43
So I think it's pretty good. It's kind of a GTFS approach of essential not exhaustive
20:22
Yep, I think you could combine that data really nicely So, you know you could street reference your curb cut or your other ramp your other feature and you could basically say like on this Street in this area is is there an accessibility feature that I need? We had a long shot with the wheel map team early on to say, you know, we don't understand all the accessibility issues
20:44
This is something that you work on. So what's important and how can we learn from this? And what they kind of said is, you know, you could combine multiple sources of data and put them together They said that as they thought about accessibility it became less of an issue about Physical infrastructure and more of just a yes-no binary. Can I get my wheelchair there?
21:03
Don't necessarily need a ramp right next to the curb Maybe something along the street would do and it really depended on the individual's mobility and what worked for them and what they were comfortable with So they kind of said it's it's a more arbitrary thing. They map businesses as yes or no. Does this work? But there's no reason why you couldn't combine that data about where the physical infrastructure is with the rule
21:26
Cool, they're telling me to kill it Thanks, everybody