Models, Gaming, and Simulation Session 8

1
Models, Gaming, and Simulation - Session 8

• The Battlefield Environment

2
Topics

• Weather and Obscuration
• Terrain Data
• Mobility Models
• High-Resolution Models
• Low-Resolution Models
• Automated Route Planning
• Combat Engineers

3
Weather and Obscuration

• Obscuration due to fog, dust, smoke or heat are
  important factors in target acquisition and
  firing accuracy.
• High- and Low-resolution models will use very
  different algorithms to account for weather and
  obscuration.
• Weather may cause trafficability and operational
  tempo effects also.
• Nuclear, Biological, and Chemical models also use
  weather sub-models

4
Weather and Obscuration

• A high-resolution approach
• Electro-Optical Systems Atmospheric Effects
  Library (E-O SAEL) maintains high-resolution
  computer models of atmospheric effects and data
• Models and Databases provide ways to represent
  local obscuration effects
• Example
• artillery smoke round impacts
• smoke-cloud software object is created
• E-O SAEL cloud growth model represents how cloud
  changes and diffuses
• target acquisition module using NVEOL model
  checks its line-of-sight against all local cloud
  objects, degrading acquisition appropriately
• smoke-cloud object is deleted when it disperses.

5
Weather and Obscuration

• A lower-resolution approach
• 5km square grid underlies terrain
• weather effects in a square are homogeneous
• weather is updated periodically in each square,
  considering
• prevailing winds
• temperature
• previous conditions
• new effects (e.g., smoke, chemicals, nuclear
  fallout, changes in weather, etc)
• Average line-of-sight and detection rate
  functions are degraded according to average
  obscuration of square(s) through which
  line-of-sight is considered.

6
Terrain Data

• Produced by National Geospatial-Intelligence
  Agency (NGA), formerly NIMA
• Three basic products (among others)
• Elevation Data
• DTED I nominally 100m postings (at equator)
• DTED II nominally 30m postings (at equator)
• DTED III nominally 10m postings (at equator)
• DTED IV nominally 3m postings (at equator)
• DTED V nominally 1m postings (at equator)
• Feature Data vegetation, cultural features,
  etc.
• Raster Data images of terrain

7
What Good is Higher-fidelity Terrain?
DTED Level I II
III IV V
8
Mobility Models

• High-resolution Models (how do vehicles move?)
• Explicit grid models
• Explicit patch models
• Implicit mobility models
• Network models
• Low-resolution Models (how do units move?)
• Tabletop models
• Regular grid models
• square
• hexagonal
• Network Models
• Combat Sector ("Piston") models

9
Explicit High-Res Mobility Models

• Explicit Grid Models
• CONCEPT maintain a multi-dimensional array of
  limiting speeds, indexed by location, vehicle
  type, vegetation, soil type, tree-stem width and
  spacing, etc. Look up the value at run time to
  determine vehicle speed.
• Problem The array must be filled, an often
  difficult task.
• Alternate CONCEPT compute limiting speed at
  run-time
• Problem Very time-consuming during model
  execution
• Explicit Patch Models
• CONCEPT compute regions which have similar
  limiting speeds store the regions as polygons
  with associated speeds. These polygons can be
  fairly large compared to vehicle size (e.g., 100
  to 500 meters in diameter), reducing storage
  requirements and look-up time.
• Example "Condensed Army Mobility Model" (CAMMS),
  a standardized U.S. Army Topographic Engineering
  model, which has evolved into the Nato Reference
  Mobility Model.

10
Example Explicit Patch Model(output from NATO
Reference Mobility Model NRMM)

11
High-Resolution Mobility Models

• Implicit models
• CONCEPT terrain is partitioned into cells, with
  average mobility factors stored in each cell.
• Advantage Compact storage and fast run-speed
• Disadvantage Does not represent the details of
  local terrain
• Network Models
• CONCEPT movement is limited to edges of a
  network
• Edges may represent roads or mobility corridors
• Advantage Compact storage and fast run-speed
• Advantage Automated route-planning algorithms
  are available
• Disadvantage Vehicle movement is not well
  represented if they are constrained to stay on a
  network edge.

12
Low-Resolution Mobility Models

• Tabletop models
• Terrain characteristics do not limit movement
• Movement rates are homogeneous, or depend on unit
  type or other factors not related to the unit's
  location
• Usually associated with older models
• Advantage Very fast
• Regular grid models (square or hexagonal)
• Each hex has mobility, line-of-sight, cover and
  concealment characteristics associated with it.
• Linear barriers to movement (e.g., rivers,
  minefields, ravines) are associated with edges
• Roads connect hex centers
• Advantages fast, allows for tactical dispersion
  of units, allows for automated route finding
• Disadvantage associated with wargame hobby
  intellectual snobbery often unfairly discounts
  the utility of hex-based models

13
Combat Sector ("Piston") Models

• CONCEPT divide the battlefield into sectors, one
  per unit. Aggregate the terrain characteristics
  in each sector
• Usually associated with theater-level models
  whose primary MOE are often related to movement
  of the FEBA (Forward Edge of the Battle Area),
  and where resources, not maneuver is the primary
  element.
• Disadvantage limits maneuver and non-linear
  battle
• Advantage may be appropriate to the scope being
  modeled.

14
Example Sector (Piston) Terrain Model

• Combat Evaluation Model (CEM) Theater Level Model
• Laydown Forces and Sectors by Unit.

15
Example Sector (Piston) Terrain Model

• Combat Evaluation Model (CEM) Theater Level Model
• Partition battlefield by each sector boundary.

16
Example Sector (Piston) Terrain Model

• Combat Evaluation Model (CEM)Theater Level Model
• Assess attrition and resulting FLOT movement by
  sector.

FLOT Forward Line of Troops
17
Network Models

• CONCEPT allow movement only centered on edges
  (like a road network)
• Advantages conforms to the barrier/mobility-corri
  dor approach to terrain analysis can use network
  search algorithms for route planning
• Disadvantages can be restrictive in some
  implementations.

18
Eagle Terrain

• Aggregate Terrain for Aggregate Units
• Two Layers of Aggregate Terrain Data
• Network of Mobility Corridors to support route
  planning and movement of military units a
  mobility corridor contains
• Trafficability data
• Average Line-of-Sight data
• Cover and concealment data
• Maneuver space data
• Underlying grid (4km x 4km) of Terrain
  Aggregates (or terags) largely for bookkeeping
• Unit occupancy lists by terag for efficiency.
• Terags also have trafficability and line-of-sight
  data to support off-corridor movement .

19
Producing Eagle Terrain
Step 1 Merge DTED Level I and DFAD data to
produce 100m x 100m cells of elevation and ground
conditions.
Step 2 For each 100m x 100m cell, compute
trafficability (one of GO, SLO-GO, NO-GO)
Step 3 Glob the data into two sets group
NO-GO cells, and group GO/SLO-GO cells discard
outliers.
Step 6 Compute trafficability and avg cover and
concealment as functions of percent GO vs SLO-GO
cells contained in each mobility corridor
compute avg and min width. Discard polygon and
cell data.
Step 5 Construct Voronoi Diagram, ie, find
center lines between NO-GO polygon edges.
Step 4 Polygonize the NO-GO globs
20
Unit Route Planning Over Corridor Network

• Unit plans route by finding optimal sequence of
  edges from Start (S) to Goal (G)
• Optimality can be in terms of shortest path,
  fastest path, safest path, or a weighted
  combination of the three metrics.
• Eagle uses A Search algorithm to find such an
  optimal route.
• This capability allows dynamic (non-scripted)
  reaction to changing events by the resolution
  units in a scenario.

21
Example Network Terrain Model

• EXAMPLE Eagle unit moves over mobility corridor
  network.

22
Automated Route Planning

• CONCEPT Provide an algorithm by which units can
  automatically find their own routes.
• This allows the analyst to focus on higher issues
  such as the overall scheme of maneuver
• This lessens the intrusion of the analyst into
  command and control, especially synchronization
  of the forces if C2 is modeled explicitly to
  allow mission-driven planning
• Units can still be given explicit routes if
  desired, or closely grouped intermediate
  objectives
• Algorithms based on graph theory
• Could be a satisficing algorithm (not guaranteed
  to find an optimal route)
• Might be an optimal algorithm
• "Optimal" may mean fastest, or shortest, or
  safest, etc.

23
Example Search Algorithm A ("A-star")

• CONCEPT from the start point work your way
  toward the goal along several alternate routes
  one step at a time. At each step, "expand" the
  search toward the node which shows the most
  promise (minimize cost so far plus estimated
  remaining distance).
• Requires a way of estimating remaining cost which
  is a lower bound on actual cost (usually
  straight-line distance)

24
A Search Example

25
A Search Example

26
A Search Example

27
A Search Example

28
A Search Example

29

A Search Example
30
Modeling Combat Engineers

• Engineers contribute to battle by modifying the
  terrain to give an advantage to the rest of the
  force. They contribute only secondarily by
  fighting.
• Therefore, modeling combat engineers will involve
  modifying the terrain.

31
Modeling Combat Engineers

• What do Engineers do?
• Mobility
• build and repair roads
• clear airfields, landing zones, FAARPS, etc
• breach barriers
• Counter-mobility
• emplace obstacles
• Survivability
• dig fighting positions and bunkers
• reinforce fixed installations
• Sustainment
• build and repair roads
• build and repair bridges
• build and improve fixed installations
• Topographic Engineering
• mapping surveying

32
Example - Engineers in Eagle

33

• As 1st Battalion moves along its route on
  mobility corridors, it may encounter a minefield
  or other obstacle (a s/w object placed on a
  mobility corridor object).
• If it knows about the minefield ahead of time,
  it would avoid attrition if not, attrition from
  mines is assessed.
• If others on 1st Bns side have already reconned
  the minefield, bypasses and breaches (objects on
  the obstacle) may be known if so, 1st Bn changes
  formation and uses one of them if not, 1st Bn
  halts in place.

34

• If 1st Battalion halts in front of the obstacle,
  it may be fired upon by enemy units (direct or
  indirect) in a vulnerable formation.
• If 1st Bn has engineer assets in its
  organization that can breach the obstacle, it
  begins to do the engineer work (implicitly, waits
  a time-delay until completed)
• If 1st Bn does not have adequate engineer
  assets, it requests engineer support from higher.

35

• The first echelon that has a supporting engineer
  unit tasks that unit to do the required engineer
  work
• The engineer unit gathers adequate assets to do
  the work and forms an engineer work team, a
  fully-qualified Eagle unit, which finds a route
  and moves to the work site, subject to normal
  attrition itself.
• When the Engineer Work Team arrives, it goes
  through the same process as the maneuver unit.

36

• The Engineer Work Team works the
  data-specified amount of time to breach the
  obstacle.
• A breach object is placed on the obstacle
  object.
• The maneuver unit finds the breach, changes
  formation, and crosses the obstacle, continuing
  its movement on the far side.
• The Engineer Work Team returns to be absorbed
  back into its parent unit.