Transcript ABM VIZ - the Atlanta Regional Commission

ARC ABM Visualization & Reporting
Activity-Based
Model
(Java, Cube)
Database
(SQL Server)
Custom
Analysis
Data Access
Layer
(IIS, ASP.Net)
Visualization
Dashboard
(Flash)
Reports
(Excel)
ARC – Nov 12, 2010
Project Need / Goal
 ABM produces a lot of data,
including new types of data, so we
need an innovative analysis tool
 Develop an ABM visualization and
reporting system
We need a system that allows us to…
Overview first, zoom and filter,
then details-on-demand
Shneiderman’s (1996) Visual Information-Seeking Mantra
Required Visuals and Reports
 Aggregate summaries
 Existing trip-based model summaries (VMT,
mode share, boardings, etc)
 New summaries (toll markets, time-of-day
results, etc)
 Geographic summaries
 County, Super District, Activity Center, Zone
 Dynamic summaries
 Maps, charts, animations, etc
 Spreadsheets
 Automatically create Excel workbooks
 Then used by ARC staff for custom analysis
 Scenario comparison functionality
ABM Visualization & Reporting System
Activity-Based
Model
(Java, Cube)
Database
(SQL Server)
Custom
Analysis
Data Access
Layer
(IIS, ASP.Net)
Visualization
Dashboard
(Flash)
Reports
(Excel)
Data Preparation
 Create De-Normalized Reporting Tables that are
FAST and SIMPLE




Add unique IDs for trips
Join tour, HH, person, skim data
Expand joint trips
Add trip start and end information, etc
 Implementation Tools
 Cube (for exporting data, networks, etc)
 Java (as part of CT-RAMP)
 SQL
 Outputs Loaded into DB
 Use BULK INSERT SQL statements
Database

Stores ABM Output Tables, Travel Time Skim Data, and
Some Inputs By Scenario
 ~8 GBs per scenario
 Around 5–10 scenarios at a time in the DB
 Can load survey data into DB as a scenario for comparison
purposes

Needs to be fast – not a traditional transactional DB,
more a data warehouse
 Use de-normalized data to avoid millions of joins each time for
dynamic visualization dashboard
 Read-only; minimal rollback features required
 Parallel query functionality (SQL Server)
 Needs excellent query caching functionality

Use ARC Enterprise SQL Server with 4 Licensed
Processors
Database Tables per Scenario
Table
Table Name
Description
Object
Primary Key
Forecast Households
FORECASTHH
Input synthesized household table
Household
HHID
Forecast Person
FORECASTPERSON Input synthesized person table
Person
PERID
Zone Data
ZONEDATA
Input zone data such as employment, county code
Zone
ZONE
Accessibility
ACCESSIBILITY
Calculated accessibility measures by zone
Zone
TAZ
Percent Walk to Transit
PCTWKT
Calculated percent of a zone that is either short walk or long walk to Zone
transit
TAZ
Household Data
HHDATA
Model results for household level choice models
Household
HH_ID
Person Data
PERSONDATA
Model results for person level choice models
Person
PERSON_ID
Work School Location
WSLOCATION
Model results for the usual work and school location (destination)
choice models
Person
PERSONID
Individual Tours
INDIVTOURS
Modeled individual tours
Tour
HH_ID, PERSON_ID,
TOUR_CATEGORY,
TOUR_ID, TOUR_PURPOSE
Joint Tours
JOINTTOURS
Modeled joint tours
Tour
HH_ID, TOUR_CATEGORY,
TOUR_ID, TOUR_PURPOSE
Individual Trips
INDIVTRIPS
Modeled individual trips
Trip
HH_ID, PERSON_ID,
TOUR_CATEGORY,
TOUR_ID, TOUR_PURPOSE,
INBOUND, STOP_ID
Joint Trips
JOINTTRIPS
Modeled joint trips
Trip
HH_ID, TOUR_CATEGORY,
TOUR_ID, TOUR_PURPOSE,
INBOUND, STOP_ID
Trips
TRIPS
Modeled individual trips and joint trips in expanded form. Also
contains joined household, person, and travel time fields.
Trip
PERSON_ID,
TOUR_ID_UNIQ, TRIP_ID
Travel Times
TRAVELTIMES
Travel times by time-of-day, and mode
OD
ORIG_TAZ, DEST_TAZ
Links
LINKS
Input and calculated link attributes by time-of-day
Link
A, B
Transit Routes
TRANSITROUTES
Transit route attributes such peak headway and daily boardings
Route
LASTNAME
Data Access Layer

Thin client to connect dynamic
visualization dashboard (Flash) and
database (SQL Server)

Use WebORB for .NET to allow Flash to
interact with the DB

Fast data communication since data
remains in binary form

Flash  .NET (C#)  SQL Server
Dynamic Visualization Dashboard
 Started with ODOT Statewide Integrated
Model Visualization Dashboard (SWIMVIZ)
 Add additional query capabilities
 Add new reports
 Excel reports
 Add new visuals
 Time Use (NY Times)
 Activity/Tour Tracing
 Radar chart
ABM VIZ – Queries



Lots of default queries
Query builder
Can write custom queries as well
ABM VIZ – Tables



Results for Queries
Results automatically spatially aggregated
(to districts, counties, etc)
Save tables to clipboard, Excel
ABM VIZ – Plots and Maps



Default queries automatically visualized
Charts linked to the map
Maps are interactive (zoom in/out, change colors, change
color bins, animate changes, etc)
ABM VIZ – Plots and Maps


Can click map zones and run current query
The above shows trip destinations for all SOV Free trips
from zones 1-10
ABM VIZ – Plots and Maps

User can change the mapped field in real time as well
ABM VIZ – Plots and Maps

Can click map zones and run current query and generate
desire lines (OD information)
ABM VIZ – Plots and Maps


Can draw the network by line width and line color
The above shows PM link volumes by PM V/C ratio
ABM VIZ – Time Use


New time use (person activity over the day)
Can select different person types (the above is showing
Full-time workers)
ABM VIZ – Time Use

The above shows Part-time workers
ABM VIZ – Tour Tracing

Can trace tours for individual micro-simulated persons
ABM VIZ – Tour Tracing
ABM VIZ – Tour Tracing
ABM VIZ – Tour Tracing
ABM VIZ – Tour Tracing
ABM VIZ – Reports




Copy tables to clipboard
Save tables to Excel
Save queries for reuse
Save maps/plots to file
Radar Chart (In Progress)
Conclusions
 Database for easier and quicker reporting
 Dashboard for real-time model results
analysis, debugging, investigation
 Platform for innovative ABM visualization
 More visuals to come
Questions
Ben Stabler, [email protected]
Joel Freedman, [email protected]