Unleash your map data: Cloud computing for geospatial applications

Geo – Tom Manshreck

The Google Maps API made geospatial development accessible to all but hosting your data remains complex and time consuming. This session will detail the services Google offers for storing your geospatial data in the cloud, illustrate the ways in which that data can be accessed and visualized, and walk through development of a retail store finder using these technologies.

Session type: 101
Attendee requirements: Experience developing web based applications that search over or visualise geospatial data is helpful but not essential.
Tags: Maps, Maps API, Data, Cloud, Geospatial
Hashtag: #geo4

Date: Wednesday May 19
Time: 4:15pm-5:15pm
Room: 4

Live Notes:

View live notes and ask questions about this session on Google Wave http://bit.ly/waveio

Tell us what you think of the I/O sessions: http://haveasec.com/io

Starting in a few minutes…

Speaker: Tom Manshreck

Maps Data API

How you can use Maps Data API to create a mashup

Private:

• My Hikes
• My Favorite Coffee Shops
• Shared
• Bike Trails
• Restaurants
• Public
• Store Finders,etc

Why store geospatial data in the cloud?

• Safer
• Exposed via multiple interfaces
• My Maps UI
• Maps Data API
• Accessible from any browser/app/server
• Open standards

Maps Data API?

• Interface to Maps Data stored in Google cloud
• REST-ish
• Operations initiated through URL requests (feeds)

Feed of the Maps Data API

• Manipulate 3 entities
• Maps ( collection of data)
• Features ( datum per map)
• Access Control Lists (per map)
• Provides searches – spatial and attribute search

POST acts to create Map

Initial Map Data can be provdied as XML, CSV or KML(best to manipulate)

Basic Feature Feed

POST acts to create Feature in Map

Features are KML

• <Point>
• <LineString>
• <Polygon>

Authentication option

• AuthSub
• OAuth
• Authorizaion via ACLs
• Specify Access on default (everyone) or per user@gmail.com

Maps Data API

URLs are a pain

Use a client library instead

• Java
• Python
• JavaScript

Loading the Maps Data API: JavaScript

• Use Google Common Loader
• Use AuthSub to redirect to Google Login (checks login status)
• Have an IMG to hold cookie token (important)

Creating Map – simple example callback functions

Application Design Specification

• coffee rating mobile application
• use Maps JavaScript API V3
• Uses Geolocation
• Allows you to click and add coffee info
• Data populated using maps data API
• Implement search over are for coffee

show application example:

• want to click and add a coffee place
• asks you to login to grant access
• click and name, add options about coffee, rating

Set up Map

• using global variables, filling in with callback functions
• create info window,geocoder and map
• call geolocate function checking where it exists with lat long – good idea to have default
• add event listener

Opening the coffee data form

• Getting coffee data
• set content of HTML and position

take out form values and pass to another function

Add features

• Get the Map feed for your Map
• Get the MapEntry and feature feed
• Populate a FeatureEntry
• Insert the FeatureEntry using the Feature Feed URL

set KML content with point values in this example (lat and long) and create a placemark point

HTTP POST to map and create a feature

Searching over your data

• Use a feature feed /snippet? query
• Pass attribute parameters in a mq parameter as key/value pair
• Spatial searches are by radius or bounding box
• radius requires a lat long center
• radius expressed in meters
• optional sortby parameter

sample shown

get back an array of features

pull out arbitrary data and populate it on a map

A Case Study – IHG Hotels – more robust example

Hotel Company with 4500 hotels to keep track of

heavy lifting done by Google

gathered data about hotels (lat long, pool, internet, etc.) and pushed data out to cloud and then display in different ways

upload hotel data nightly

simple REST query to perform search and then display data in map

iPhone and Android apps available too

using Google Maps Data API – problem resolved in 24 hours; traditional way would have taken a lot longer

It’s Simple! Push your data, query it, results returned quickly, reliable

• JSON feature Upload/Download coming soon
• no need to parse KML
• perfect match for users of the JavaSCript library
• FusionTables in the JavaScript V3
• have columns and form SQL queries