Download presentation
Presentation is loading. Please wait.
Published byAusten Griffin Modified over 9 years ago
1
ARC ABM Visualization & Reporting ARC – Nov 12, 2010 Activity-Based Model (Java, Cube) Activity-Based Model (Java, Cube) Database (SQL Server) Visualization Dashboard (Flash) Visualization Dashboard (Flash) Reports (Excel) Data Access Layer (IIS, ASP.Net) Data Access Layer (IIS, ASP.Net) Custom Analysis
2
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
3
We need a system that allows us to… Overview first, zoom and filter, then details-on-demand Shneiderman’s (1996) Visual Information-Seeking Mantra
4
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
5
ABM Visualization & Reporting System Activity-Based Model (Java, Cube) Activity-Based Model (Java, Cube) Database (SQL Server) Visualization Dashboard (Flash) Visualization Dashboard (Flash) Reports (Excel) Data Access Layer (IIS, ASP.Net) Data Access Layer (IIS, ASP.Net) Custom Analysis
6
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
7
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
8
Database Tables per Scenario TableTable NameDescriptionObjectPrimary Key Forecast HouseholdsFORECASTHHInput synthesized household tableHouseholdHHID Forecast PersonFORECASTPERSONInput synthesized person tablePersonPERID Zone DataZONEDATAInput zone data such as employment, county codeZoneZONE AccessibilityACCESSIBILITYCalculated accessibility measures by zoneZoneTAZ Percent Walk to TransitPCTWKTCalculated percent of a zone that is either short walk or long walk to transit ZoneTAZ Household DataHHDATAModel results for household level choice modelsHouseholdHH_ID Person DataPERSONDATAModel results for person level choice modelsPersonPERSON_ID Work School LocationWSLOCATIONModel results for the usual work and school location (destination) choice models PersonPERSONID Individual ToursINDIVTOURSModeled individual toursTourHH_ID, PERSON_ID, TOUR_CATEGORY, TOUR_ID, TOUR_PURPOSE Joint ToursJOINTTOURSModeled joint toursTourHH_ID, TOUR_CATEGORY, TOUR_ID, TOUR_PURPOSE Individual TripsINDIVTRIPSModeled individual tripsTripHH_ID, PERSON_ID, TOUR_CATEGORY, TOUR_ID, TOUR_PURPOSE, INBOUND, STOP_ID Joint TripsJOINTTRIPSModeled joint tripsTripHH_ID, TOUR_CATEGORY, TOUR_ID, TOUR_PURPOSE, INBOUND, STOP_ID TripsTRIPSModeled individual trips and joint trips in expanded form. Also contains joined household, person, and travel time fields. TripPERSON_ID, TOUR_ID_UNIQ, TRIP_ID Travel TimesTRAVELTIMESTravel times by time-of-day, and modeODORIG_TAZ, DEST_TAZ LinksLINKSInput and calculated link attributes by time-of-dayLinkA, B Transit RoutesTRANSITROUTESTransit route attributes such peak headway and daily boardingsRouteLASTNAME
9
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
10
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
11
ABM VIZ – Queries Lots of default queries Query builder Can write custom queries as well
12
ABM VIZ – Tables Results for Queries Results automatically spatially aggregated (to districts, counties, etc) Save tables to clipboard, Excel
13
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)
14
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
15
ABM VIZ – Plots and Maps User can change the mapped field in real time as well
16
ABM VIZ – Plots and Maps Can click map zones and run current query and generate desire lines (OD information)
17
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
18
ABM VIZ – Time Use New time use (person activity over the day) Can select different person types (the above is showing Full-time workers)
19
ABM VIZ – Time Use The above shows Part-time workers
20
ABM VIZ – Tour Tracing Can trace tours for individual micro-simulated persons
21
ABM VIZ – Tour Tracing
25
Copy tables to clipboard Save tables to Excel Save queries for reuse Save maps/plots to file ABM VIZ – Reports
26
Radar Chart (In Progress)
27
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
28
Questions Ben Stabler, stabler@pbworld.comstabler@pbworld.com Joel Freedman, freedman@pbworld.comfreedman@pbworld.com
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.