Merken

Developing Tools for Humanitarian Decision Making

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
but the there we go ahead but about the so that now we have a lot of them so the the the sort Excel silos over the organization and basically what happened was a created calls richest people not talk and other people are not even sort of do not to like I mean sort whatever you know you had your working you did you work and other people could take advantage of it was and this is not a story meets the red cross right so to so a long time ago 17 years ago to to backtrack a little bit as well as the persons are people message you
know we need we need a system we need a system that is going to allow us to sort of in a pool of all ideas yeah there
and it's going to take the magical right but I and I sort of it's almost like the stock go machine year of like it's the it just doesn't work like that it can't pull pieced together it took a long time to try and
work it ultimately failed a bunch and I
saw this 1 from phosphor G website so for
for and but it's if it didn't work
right the organization was ready technology was a
ready I and then so are a lot of effort
went into defining sort of how manager did
look better and so they arrived at a cost and sales force application is what the coast which is efficient collaboration collaborative operate system we love our happens at the repressed the only problem is sales force by default it's not spatial so we have all this data now we're putting it into the system but we couldn't actually you could see on map we know where stuff was you know what what you know where your product and he where hate and so that the sine die I hired by the across and so my 1st job
my force a job I was told you need to make cells for spatial need to build a custom web at that
map application that can just see this
so we can look at the at this time there's also at
history of of companies
that were hired by the across that had
I that shall remain nameless that had sort
of a more proprietary beds on how to do
this and Chai said invasive favor the context in
work they sort of what kind enough our
customer focus for us is to have missed the mark
completely so knowing this year we
had a lot of defer requirements lot technology and 1 it's really sort of bold things up
I decided to hire 2 people a lot smarter than me to put out an open bed immediate hiring spatial that in Seattle it's Ryan to board makers and the really nice guys of I ran we met them at JHU I think a couple years ago I'm impressed you some stuff they're doing in an the so we have we really just didn't we didn't want a typical search consultant I wouldn't want to deceive the beltway bandits of which there are many in the number of firms that will gladly take your money bills to assist and walk away and sensitive sort of where are history of it of starts and stops in the mall books a died the application of so 1st thing we had to do was we had to but yeah figure out how to what happened at the 1st thing we need to do all of the stuff we're working through this out of the 1st things we did do was figure had tag non-spatial project locations that people at right cross-word entry into the cell source application with regret tagged this projects with locations of and the way that we decided to do that was tagging and to political
boundaries rather than lack village
fantasy so 1st thing to do
was figure out what boundary dataset used the and I said that if that yes so we do we we try to use a bunch of boundaries and basically and call doesn't work this license is bad under like admin 3 in some places datum design admin 5 only 2 countries and we're assertive OSN is a great database was an admin 10 but it's not use instead set the work with on the China plots of 1 boundary right so we always problems so we decided well this thing is example please cite a sort
of figure out always use built sorry new structure gets yet so that
so in doing using the call us over the gas phase the standard that goes down to level 5 in some countries so if enough the bubbles zeros country level 1 is equal to the states of all 2 counties so when somebody at Red Cross 1 tag where project disasters happening they could go to a Web service that we build a little web form of start typing a place name you to place name search and search through all of the academy boundaries and then return result that we would store that ID that we knew later on we're going to be building a web app and 1st step here was just a the tag
locations so after that was done in the next thing we need to do which was the previous slide
actually which was set up parent-child
relationships which was important things was on the map application being a role by being able roll up the the number of projects the number of disasters happening in different areas in order to do that we need to know you know if you a a disaster Apple Mobile County for example then when you resume that to United States we needed to be able to represent that number up to that bubble about to country-level so being able to serve this parent-child relationships which is of this slide is about I was important so we're able to do some preprocessing and some aggregation before and so we're tagging project locations with of boundaries and at every level bubbling up and in the next slide in order to serve this stuff out and thankfully coincidentally I've been working on a node j s application that serves out
duties on from unstressed and post yeah approach of observer and the slide jobs so and thankfully that was the requirement that have that I wanna know journalists PostGIS applications of great have been working on that so so if you go the next 1 so we're starting to
get some pieces of that node j is application
we've got about boundary dataset with the project
locations in because now we start to
pull this together of getting the data will
be getting the data out of sales force was another
component worked out so we wrote a custom no JS ETL to reach into cells forces API extract just the data
that we're working with here all kinds of data there too
crazy data model really put the pieces that we needed basically the projects and
disasters and associated details so after pull that out we're able to then
Munger together after with our boundaries and having in our own POS press
instance we can really manipulated so that's
representing the 2 so that this point we've got all the data that match together it's ready to be shown just waiting and that the the next from we tackle that was figuring out when to show what boundaries where and at what point so for example if you're looking at at a web map as somebody's tagger a project County then would you do if you assume that to the country GID showing the USA boundaries the state boundaries in that state solution county vendors have you know what to show with the remember we were with G some we don't wanna show all the states or all of the counties in the US on point at 1 time even in in the viewport looking at so we came up with this something we call a smart and what is basically a relatively simple post press function that simple gonna time us Moghaddam rules related to that the nodes of the kind of so so again base there's is a function that decides what to display
the user based on how many features of looking at so for example as you're browsing around map if you see uh 3 or
fewer countries we actually remove the countries and show the next down which states if you keeps them again if you showing 3 or fewer states that we turn that often turn on counties helping user and of
obviously have a manual override so if the user wants to see a level 0 they can force on the whole here a little too but the idea here is being able to sort of know interchange between the levels and habitat and being able to show users different levels of detail about that so thought about there's multiples of boundary issues here again this is not an issue that out there for together project I think as a start getting into we're surprised the complexity of this problem this shows that yeah basically so Dad has as we discovered when we initially you simplify the geometry using it to simplify posters to make a render on the level of a faster in terms rigorous application we use the all these little blue dots on from basically that the topology of down wrong there are with 2 the global datasets there are 200 something doesn't anthropology topology areas and gas and which doesn't work right the starts of the 5 and strategies using you start with the goal of all there's in France universe was this is just this admin alright so that we know that we used a very complex process 1 of our wonderful inference figure out and solve the problem of the it was on there I will post
the itself but we don't want to everything you you did it work
so this get out of common across
the the aggressive we if we basically it
to make it work but in the end this works in the application but not the got the summary so having problems that we gave up on them because of this India Indonesia still very large it too long will on in matter so gave up on judicial together into vector tiles in application and this is where we we thought we figured out in a city girl here you go guys and they're wired up was that the going was basically is out of the cell but we have people who look up here this is the of the rich
so here we have the titans of India the search is that this is no 0 himself was the tiles review for the of the and you can't get the size of the use of 3 here this find the whole stack if you will for that area which will work for us for aggregating down all of the countries continents regions counties and prefectures whatever your favorite book the the since you brought up all modular quickly so that the expected tells is actually map
make PPF proud above vector
tiles of being consumed and rendered
in the Flickr which critical so that's what that was huge how no problem labels labels
so yes of so 1 of the things we could
think about was let's say semantic tag
that there was a disaster happening in India where
we were you put the label for that on the map you
know let's say we want to label the number of
disasters happening in the
world you could very
pre-calculate the centroid of India and store
the database and then use that as a label what were
what have somebody that is the map over so that the
central minutes after that we still want to
display a number of still wants display some
for the users see so make them here at the
front door created a leaflet
plug-in that loop is dynamically
generating centroids based on the
visible area of the features looking at or
features that are that are visible screen so for
example if you're only looking at half of India
and wikis dynamically calculating on
every day and every pair the centroid
of just the visible area India so that
even if you planning in in your only looking at a little
sliver of India you still get a center is
centroid of that is dynamic calculated is
also we love it and it means you never without
a label this great as long as you're looking at a piece of
the geometry through like this is the stack
so sort of Tigon altogether too soon to tell you a little bit about false that is 0 which post present back and for spatial data is is that each of its Poland in that sales force matching up with that spatial data and on top of that we've no JS apt culture the spatial server which is again serving out the duties on a can do dynamic querying serving up the details as to local stuff we also of custom and points in there to act is cast as web services for for the actor Richard Secord Amendment up about reusing leaflet and and you're in front of the rendering from the
world to show that the
a I this is actually a
so here see applications of your or across regions how we divide the world you have the Middle East in Europe so largest 1 basically have world where the world view of the collapse of last of the programs and here we see the labels actually 1 of the initial probably have finally have that the with the same thing we're going and it was too much land mass of the label that was the way all of being in Canada TFs used to modulate the breakthrough of Canada from the dataset for now we're at war with the humans you Africa using the program will
stop here of the Johnny click on it including different types of programs here the the the breakdown the difficult 1 of united programs cells see more data about it and it's even more data
from those will try to really try to sort of solving we're not replicating the sales
force because you never and there's so many but some data that's available right we're trying to make it so that the information that is something I make 5 minutes is or get that to the whole operating on the goes we also through the the blood the unity Mori Wyatt stuff that came back out the sales force on this so
that those words we have if you
want to think the project but let's look at the answers but the whole
the red here the west African people operate the the long the fact that you have disaster the more information about the world we also you stories
this is really before us work as we as a team of it's just is not all of the size of a lot of sort of self service a lot of of data on on the needs for mass so this idea of stories that follows the culture that people could create a roadmap and the Euler the the regular
actually this is the really great that obviously and somebody else not run at all it will be there so that these people this is the last the 3 shots and you can upload that and see the story you will have to know but has no fancy red so as to build up build up the application In the future of the hundreds of stores thousands of stories you can find what you want and this is often because you can click on a story and using angular exactly replicate
so you saw the history it also
so we want to look at the health so
recessive results a risk that
our the for world like you know this there is this the
basically the residence note from Figure 1
of things that the the part of the so
you get the the how well the properties
are of we of the complex are actually but I'm a bunch of different metrics so things that made metrics of stopwords schema is produced and at the that you know the contract medical 14 months so maybe the party has 4 words that are usually going out 1 of the contractors was was over budget or maybe on both of you know it's all different if the particles so this is the fact allowing users to see very briefly what's going on well this
is the year of the apparently what I want is crucial adding layers all to how we would approach approaches limit differently was that we did not want to but we now want to have a thousand layers that you so that the user is the have I mean it is sort of the dual is attached at the end of the the will change I tried to solve the problem of you know put into a treaty quickly into another treaty it's video you wanted and then gene something and then you for our unchecked balance so that when you 2 layers and whatever so we also saw that as well we always search for doing so on at the end it so as people that do is time of a population that's really the statement of the system and its search for the attitude of will also possible and all the your data from and European uh just duties on Kim elasticity very wrote the on
just online so it's not on our servers so also gets to the not someone else say so you know will be a the to well so that was thought so you sort of access but the all right so we have different worlds with the that from the the people projects so did the this of status and of provided this is actually doing it this functionality works with the back datasets supported which have emerged in the fact that the the and need the some power users everyone mucking around here to queries the concerns of the query using this stopping distance somebody else and that they can the same query and using vision is for they offered by initially but have heard some good things from from the this operations financial voltages of more cost to cooperation the legend reasonable for now you want
to watch pile of based maps the every company be but the
analysis yeah geographic resource
FIL awareness and but you data actually did the good on this
allows us to of service allow the use of freedom that date it is that in a lot of service right so if they need to create a map with that on the 1 or more of so back the background there 1 all right so they 1 that will
work with the you know that you have these
that allowing users of such a lot more work firstly like coding of of a have not in that the problem is that but allows them to search selection and time into it and we also the next work here and you can save your else's somebody in so if you're worried about the powerpoint or document to like the back on the Net using phantom debts over thing right we also have search so you want for
project the I you on the slides
you again a word find a strip away all the in knowledge traditional due stuff on initial or talk about what is that this is like moving forward and in other people's banana across a which at this which that that I I think the myself instead of the switch guys were derelict of pushing back against that of that no it doesn't actually need a truck feature you knew were 3 data in this in this matter what's going on at least these this to that black is for 1 purpose as opposed to 1 application of many purposes and then you serve your purpose is modeled as the reason for the application becomes low value instead of that you wanna stay the the courts what was and everything that that people think that you something where the
figures like challenged paradigm in
organizations challenge sort of higher applications
about the original scope of the project was
to use them as you database and use ArcGIS
Online and we but that's a challenge of
what people want consultants to tell you that
you have consultants yeah consultants will
challenged the paradigm in your own organization
it was really about art I find achieving figured
out to me on my last which virus
so the the Jewish started with but the
black that thank you thank you
without
2 questions that he really wanted to reference
I don't think so but I want you to
take a lot of things might what
the who
the heart
and the work the we do on on the low side
projects and we are ability
of me all are so this is an
internal version of this application you know it's it's
behind the sales force here and
we need to make this data of where appropriate
available through use energy on the
a half of the 1st of
creating an application or service
allows us to how that's fair enough some
files and we're going to take use all this
technologies that we use in this application to fill the service to the query get the real time but
where what are projects are more than that
of the kernel there is that they're
this doesn't matter how about the universe and
so are that there's so because we have access
to the Assembly places that have system and
they are using sort order is so that we
will the good name of the of
the except that the organism and at this
1 of the the
NASA is the gas services about
200 people but in accordance nature and
the use of the center of the world it is use relative from hostage to the
person responsible for complimentaries was but
it's so that the director lower they're looking
at sort of each so Africa and it also
Rob surveys to the question
arises graph operations 3 passports for him so he can see exactly what it
already map about some of the mass of the knew where the weekly monthly what's going on in the
makes that could
Bit
Rechter Winkel
Selbst organisierendes System
Systemaufruf
Message-Passing
Quick-Sort
Virtuelle Maschine
Rechter Winkel
Physikalisches System
Quick-Sort
Computeranimation
Web Site
Datenmanagement
Selbst organisierendes System
Quick-Sort
Computeranimation
Mapping <Computergraphik>
Kollaboration <Informatik>
Forcing
Prozess <Informatik>
Kartesische Koordinaten
Physikalisches System
Biprodukt
Default
Mapping <Computergraphik>
Benutzerbeteiligung
Forcing
Prozess <Informatik>
Zellularer Automat
Kartesische Koordinaten
Quick-Sort
Fokalpunkt
Quick-Sort
Randwert
Rechter Winkel
Zellularer Automat
Zahlenbereich
EDV-Beratung
Kartesische Koordinaten
Projektive Ebene
Quellcode
URL
Figurierte Zahl
Quick-Sort
Whiteboard
Randwert
Rechter Winkel
Datenhaltung
Systemverwaltung
Plot <Graphische Darstellung>
Figurierte Zahl
Quick-Sort
Resultante
Web-Applikation
Systemaufruf
Übergang
Randwert
Benutzerbeteiligung
Bildschirmmaske
Web Services
Projektive Ebene
Datenstruktur
Figurierte Zahl
Aggregatzustand
Standardabweichung
Subtraktion
Präprozessor
Mobiles Internet
Zahlenbereich
Kartesische Koordinaten
Übergang
Rechenschieber
Mapping <Computergraphik>
Randwert
Knotenmenge
Flächeninhalt
Projektive Ebene
URL
Ordnung <Mathematik>
Gleitendes Mittel
Rechenschieber
Knotenmenge
Prozess <Informatik>
Luenberger-Beobachter
Kartesische Koordinaten
Randwert
Forcing
Zellularer Automat
Zusammenhängender Graph
Projektive Ebene
URL
Mapping <Computergraphik>
Lineares Funktional
Randwert
Knotenmenge
Benutzerbeteiligung
Punkt
Datenmodell
Schlussregel
Projektive Ebene
Aggregatzustand
Instantiierung
Mapping <Computergraphik>
Aggregatzustand
Inklusion <Mathematik>
Subtraktion
Prozess <Physik>
Kolmogorov-Komplexität
Inferenz <Künstliche Intelligenz>
Kartesische Koordinaten
Term
Räumliche Anordnung
Quick-Sort
Computeranimation
Übergang
Netzwerktopologie
Randwert
Multiplikation
Flächeninhalt
Einheit <Mathematik>
Strategisches Spiel
Volumenvisualisierung
Projektive Ebene
Grundraum
Figurierte Zahl
Funktion <Mathematik>
Einheit <Mathematik>
Parkettierung
Zellularer Automat
Kartesische Koordinaten
Vektorraum
Computeranimation
Mapping <Computergraphik>
Flächeninhalt
Parkettierung
Vektorraum
Dialekt
Computeranimation
Mapping <Computergraphik>
Parkettierung
Vorlesung/Konferenz
Flickr
Computeranimation
Formale Semantik
Mapping <Computergraphik>
Datensichtgerät
Datenhaltung
Zahlenbereich
Vorlesung/Konferenz
Speicher <Informatik>
Loop
Flächeninhalt
Touchscreen
Flächeninhalt
Vorlesung/Konferenz
Punkt
Web Services
Forcing
Server
Kombinatorische Gruppentheorie
Quick-Sort
Sichtenkonzept
Ruhmasse
Kartesische Koordinaten
Optimierung
Dialekt
Computeranimation
Subtraktion
Datentyp
Zellularer Automat
Optimierung
Quick-Sort
Nichtlinearer Operator
Forcing
Rechter Winkel
Wort <Informatik>
Information
Eins
Projektive Ebene
Information
Dienst <Informatik>
Ruhmasse
Kartesische Koordinaten
Speicher <Informatik>
Quick-Sort
Computeranimation
Resultante
Computeranimation
Subtraktion
Kolmogorov-Komplexität
Linienelement
Kategorie <Mathematik>
Mereologie
Wort <Informatik>
Partikelsystem
Figurierte Zahl
Computeranimation
Design by Contract
Nichtlinearer Operator
Lineares Funktional
Befehl <Informatik>
Abfrage
Euler-Winkel
Physikalisches System
Quick-Sort
Videokonferenz
Server
Projektive Ebene
Abstand
Elastische Deformation
Maschinelles Sehen
Leistung <Physik>
Mapping <Computergraphik>
Kontextbezogenes System
Analysis
Mapping <Computergraphik>
Dienst <Informatik>
Rechter Winkel
Computeranimation
Rechenschieber
Rechter Winkel
Trennschärfe <Statistik>
Projektive Ebene
Computeranimation
Wort <Informatik>
Kartesische Koordinaten
Figurierte Zahl
Selbst organisierendes System
Datenhaltung
Programmierparadigma
EDV-Beratung
Projektive Ebene
Kartesische Koordinaten
Quick-Sort
Computervirus
Projektive Ebene
Data Mining
Forcing
Versionsverwaltung
Kartesische Koordinaten
Exergie
Dienst <Informatik>
Abfrage
Projektive Ebene
Kartesische Koordinaten
Kernel <Informatik>
Ordnung <Mathematik>
Quick-Sort
Dienst <Informatik>
Natürliche Zahl
Relativitätstheorie
Nichtlinearer Operator
Graph
Endogene Variable
Ruhmasse
Sondierung
Quick-Sort
Computeranimation

Metadaten

Formale Metadaten

Titel Developing Tools for Humanitarian Decision Making
Serientitel FOSS4G 2014 Portland
Autor Kunce, Dale
Lizenz CC-Namensnennung 3.0 Deutschland:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen.
DOI 10.5446/31591
Herausgeber FOSS4G, Open Source Geospatial Foundation (OSGeo)
Erscheinungsjahr 2014
Sprache Englisch
Produzent FOSS4G
Open Source Geospatial Foundation (OSGeo)
Produktionsjahr 2014
Produktionsort Portland, Oregon, United States of America

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract The American Red Cross International Services Department (ISD) and project partners are developing a web visualization tool (Mapfolio) to help to visualize Red Cross's disaster response and humanitarian projects around the world. The solution uses innovative Node.js ETL processing to process information from the Red Cross Salesforce Information Management platform. The Mapfolio is open source built on Node.js, Angular, Leaflet, PostGIS, and a custom PGRestAPI (Chubbs). Other technical advances include a custom Leaflet clipped-polygon labeling as well as a map-view-dependent (not zoom level dependent) display of global administrative boundaries. This session will not only walk through the open source components but will also focus on how the Red Cross defined clear deliverables and scaled-up its support of free and open source software.
Schlagwörter Humanitarian Relief
Disaster response
Operations
Leaflet
Angular.js
Node.js
Postgres

Ähnliche Filme

Loading...