Easing the Pain of Urban Modeling

1
Easing the Pain of Urban Modeling

• John Mann
• Applied Research Associates, Inc.
• john.mann_at_ara.com
• Latika Bonnie Eifert
• RDECOM/STTC
• Latika.Eifert_at_us.army.mil

2
Overview

• Typical uses of urban databases
• Urban modeling challenges
• Easing The Pain
• Automated building generation
• Rapid geotypical and geospecific building
  creation
• Adding realism to building models

3
Typical Uses of Urban Databases

• Military
• Exercises
• Training individual, collective, command level
• Mission rehearsal, mission planning
• Course of Action analysis
• Experimentation
• Civil
• Emergency responder training
• City planning
• Entertainment
• Gaming
• Movies

4
Video Flyover
5
Military Applications
Large Scale Exercises
Constructive Simulation
6
Civil Applications
City Planning
Google Earth
7
Entertainment
Full Spectrum Warrior
First to Fight
8
Urban Modeling Challenges

• Producing high fidelity databases
• Building interiors, furniture, lighting
• Urban clutter lightposts, trash cans, fire
  hydrants
• Streets, sidewalks, curbs
• Fast turn-around time
• 96 hour turn-around from data collection
• Overcoming limitations of tools
• Most tools require manual intervention
• No seamless end-to-end solution exists

9
Urban and Underground Model Generator (U2MG)

• Urban modeling solutions
• Must replicate the rich detail in an urban
  environment
• Should include modeling of structural properties
• U2MG leverages over 10 years of development
  effort
• Structural knowledge
• Intelligent building creation
• Data driven approach to building generation

10
Modeling Comparison

• Modeled in Multigen Creator
• Modeling effort
• Create and position walls
• Create and position apertures
• Create and position stairs (if applicable)
• Create roof
• Create LODs
• Add textures
• Add SAF attribution
• Import model into TV and build database
• Test and fix SAF problems
• Total modeling time 4 days

• Modeled in U2MG editor
• Modeling effort
• Set parameters in building wizard
• Click Finish to create model (model is
  auto-generated)
• Manually edit building layout
• Import model into TV and build database
• Total modeling time lt 5 minutes

11
Manual vs Automated Process
GIS source data
Automated Process
Manual Process
Create and position walls and floors
Set initial building parameters
Normally takes minutes or hours to model
each building
Create and position windows and doors
Auto-generation
Architectural
Architectural
Create and position stairs (if multi-story)
Edit auto-generated layout if necessary
assessment
assessment
Normally takes days or weeks to model
each building
Create Level of Detail (LOD) nodes
Import into database
Add textures
Database errors are usually minor or non-existent
Add SAF attribution
Import into database
Validate database and fix errors
12
Automated Building Creation
Step 1 Create entrances and public areas (e.g.
hallways, elevators, staircases and bathrooms)
Step 2 Determine the remaining areas (areas left
over after subtracting public areas)
Step 3 Subdivide remaining areas into rectangles
(multiple options)
Step 4 Subdivide rectangles into offices
(maximize number of offices) and add windows and
doors
13
Building Layout Parameters

• Layout parameters are specified in an XML file
• Room types, priority, size
• Relationships between rooms
• Optional rooms
• Stairs, elevators
• Door and window size and location
• Parameters can be relative or absolute

14
Automated Geotypical Buildings
2 Elevators
2 Bathrooms
Staircase
Exit
Corner Office
Entry Door
Exit
Hotel
Main Entrance
Office Building
Staircase
2 Public Bathrooms
Exit
2 Elevators
Kitchen
Bathroom
Office
Reception
Exit
Breakfast or Sitting Area
Hotel Room (2 Queen Bed)
Hotel Room (1 King Bed)
15
Geospecific Buildings
CAD drawing of office building in downtown Orlando

• 3D model of the
• building created
• in the U2MG editor

16
Adding Realism to Models

• Buildings with balconies, porches, facades,
  courtyards
• Furniture placement
• Material properties
• Building damage

17
Conclusions

• Automating key processes is critical to fast
  turn-around of high fidelity urban databases
• General purpose solutions are not sufficient
  need a data driven solution
• Realistic training requires databases that
  replicate the physical world as closely as
  possible

18
Questions?