SCM Developer Guide

Visual Perception

The top-down process in SCM is illustrated in Fig. 68.

Fig. 68 Top-down information request mechanism

InformationAcquisition - own vehicle information

PerceiveEgoData is called by SCM in every time step, regardless of any perception states of the driver’s sensory system. It transfers the following information about the driver’s own vehicle to the submodule Mental Model, more specifically old ref: microscopic_characteristics :

Variable	Meaning
AbsoluteVelocity	Absolute velocity of the vehicle in the road coordinate system
LongitudinalAcceleration	Longitudinal acceleration of the vehicle in the road coordinate system
LongitudinalVelocity	Longitudina velocity of the vehicle in the road coordinate system
LongitudinalPosition	Longitudina position of the vehicle in the road coordinate system (current track coordinate)
LateralVelocity	Lateral velocity of the vehicle in the road coordinate system
LateralPosition	Lateral position of the vehicle in the road coordinate system
HeadingAngle	Heading angle of the vehicle to the road coordinate system
isCrossingLanes	Boolean indicating, if the driver’s own vehicle is currently attached two more than one lane (e.g. in case of a lane change)
Collision	Boolean indicating, if the driver’s own vehicle has collided with another object
laneId	ID of the driver’s current (main) lane
LeftLaneBoundaryDepartureEgo	Minimum lateral distance of the vehicle to its current lane’s left marking
RightLaneBoundaryDepartureEgo	Minimum lateral distance of the vehicle to its current lane’s left marking

InformationAcquisition - surrounding vehicle information

This information is stored/update in MentalModel for every AreaOfInterest in which an object is observed. With exception to the AreaOfInterest::LEFT_SIDE / RIGHT_SIDE / LEFTLEFT_SIDE and RIGHTRIGHT_SIDE only one object, the one closest to the ego vehicle, is saved to the driver’s mental model. For these mentioned side areas, all vehicles are stored in a vector.

Variable	Meaning
acceleration	Absolute acceleration of the object
accelerationDelta	Relative longitudinal acceleration of the surrounding object and the driver’s own vehicle in the road coordinate system
brakeLightsActive	A Boolean indicating, if the surrounding object has activated brake lights or not
collision	A Boolean indicating, if the surrounding object was involved in a collision
distanceToLaneBoundaryLeft	Minimum lateral distance of the surrounding object to its current lane’s left marking
distanceToLaneBoundaryRight	Minimum lateral distance of the surrounding object to its current lane’s right marking
exist	A Boolean indicating, if the surrounding object exists or not
gap	Time headway between the surrounding object and the driver’s own vehicle in the road coordinate system
gapDot	Time derivative of the time headway between the surrounding object and the driver’s own vehicle in the road coordinate system
heading	Heading angle of the surrounding object to the road coordinate system
height	Maximum height of the surrounding object
id	ID of the surrounding object
indicatorState	IndicatorState of the surrounding object
isCrossingLanes	A Boolean indicating, if the surrounding object is currently crossing lanes
isStatic	A Boolean indicating, if the surrounding object is a static object or a moving agent
lateralDistanceDelta	Relative lateral distance of coordinate reference points between the surrounding object and the driver’s own vehicle in the road coordinate system
lateralPositionInLane	Lateral position of the surrounding object in the road coordinate system
lateralVelocity	Lateral velocity of the surrounding object in the road coordinate system
lateralVelocityDelta	Relative lateral velocity between the surrounding object and the driver’s own vehicle in the road coordinate system
length	Maximum length of the surrounding object
longitudinalObstruction	Obstruction between the object and ego
longitudinalVelocity	Longitudinal velocity of the surrounding object in the road coordinate system
mainLaneId	Main lane ID of the surrounding object
relativeLateralDistance	Relative lateral net distance between the surrounding object and the driver’s own vehicle in the road coordinate system
relativeLongitudinalDistance	Relative longitudinal net-distance of the surrounding object to the driver’s own vehicle in the road coordinate system
tauDot	Time derivative of the time to collision between the surrounding object and the driver’s own vehicle in the road coordinate system
ttc	Time to collision between the surrounding object and the driver’s own vehicle in the road coordinate system
velocity	Absolute velocity of the surrounding object in the road coordinate system
velocityDelta	Relative longitudinal velocity between the surrounding object and the driver’s own vehicle in the road coordinate system
width	Maximum width of the surrounding object

Fig. 69 Calling sequence for the visual perception of microscopic information from an AreaOfInterest

For AOIs in which no vehicle is observed, the values of the struct are reset to the following:

Variable	Dummy value for empty AreaOfInterest
acceleration	0.0
accelerationDelta	0.0
brakeLightsActive	false
collision	false
distanceToLaneBoundaryLeft	0.5
distanceToLaneBoundaryRight	0.5
exist	false
gap	PreviewDistance of the driver divided by the longitudinal velocity of the driver’s own vehicle (the denominator is bound to 1.0 below)
heading	0.0
height	0.0
id	-1
indicatorState	IndicatorState_Off
isCrossingLanes	false
isStatic	false
lateralDistanceDelta	PreviewDistance of the driver
lateralPositionInLane	0.0
lateralVelocity	0.0
lateralVelocityDelta	0.0
length	0.0
longitudinalObstruction	NoOpponent
longitudinalVelocity	0.0
mainLaneId	0
relativeLateralDistance	PreviewDistance of the driver
relativeLongitudinalDistance	PreviewDistance of the driver
tauDot	0.0
ttc	99.0
velocity	0.0
velocityDelta	0.0
width	0.0

The following OpticalData information is corrected by the function AssignAoiSpecificDataForObject, if the specified AreaOfInterest is one of the SIDE areas:

Variable	Correction for AreaOfInterest LEFT_SIDE or RIGHT_SIDE
DeltaLongitudinalAcceleration	0.0
DeltaLongitudinalVelocity	0.0
RelativeNetDistance	0.0
DeltaLateralVelocity	0.0
Gap	0.0
GapDot	0.0
Ttc	99.0 s
TauDot	0.0

InformationAcquisition - information about infrastructure

The information from struct TrafficRuleInformationSCM consist of the information shown below (see InfrastructureInformation in SensorDriver).

Variable	Meaning
laneLeftLeft - trafficSigns - laneMarkingsLeft - laneMarkingsRight	Data about the lane two lanes left of the mainLane (in driving direction) - All traffic signs on this lane inside the visibility distance - Lane markings on the left side of the lane - Lane markings on the right side of the lane
laneLeft - trafficSigns - laneMarkingsLeft - laneMarkingsRight	Data about the lane to the left of the mainLane (in driving direction) - All traffic signs on this lane inside the visibility distance - Lane markings on the left side of the lane - Lane markings on the right side of the lane
laneEgo - trafficSigns - laneMarkingsLeft - laneMarkingsRight	Data about the lane where the middle of the front of the agent is (e.g. mainLane) - All traffic signs on this lane inside the visibility distance - Lane markings on the left side of the lane - Lane markings on the right side of the lane
laneRight - trafficSigns - laneMarkingsLeft - laneMarkingsRight	Data about the lane to the right of the mainLane (in driving direction) - All traffic signs on this lane inside the visibility distance - Lane markings on the left side of the lane - Lane markings on the right side of the lane
laneRightRight - trafficSigns - laneMarkingsLeft - laneMarkingsRight	Data about the lane two lanes right of the mainLane (in driving direction) - All traffic signs on this lane inside the visibility distance - Lane markings on the left side of the lane - Lane markings on the right side of the lane

The information of struct GeometryInformationSCM is composed as follows (see InfrastructureInformation) in SensorDriver

Variable	Meaning
visibilityDistance	Current maximum visibility distance as specified by the world
numberOfLanes	Number of lane at the current position
distanceToEndOfNextExit	Distance from current position to the end of the next (current or coming) exit lane
distanceToEndOfNextEntry	Distance from current position to the end of the next (current or coming) entry lane
laneLeftLeft - exist - curvature - width - distanceToEndOfLane - curvatureInPreviewDistance - curvatureNextPosition	Data about the lane two lanes left of the mainLane (in driving direction) - Wether there is a lane on this position - Curvature at current s position - Width at current s position - Distance from current position to the end of the lane - Curvature at previewDistance - Curvature at position the agent will be in the next time step (it constant velocity)
laneLeft - exist - curvature - width - distanceToEndOfLane - curvatureInPreviewDistance - curvatureNextPosition	Data about the lane to the left of the mainLane (in driving direction) - Wether there is a lane on this position - Curvature at current s position - Width at current s position - Distance from current position to the end of the lane - Curvature at previewDistance - Curvature at position the agent will be in the next time step (it constant velocity)
laneEgo - exist - curvature - width - distanceToEndOfLane - curvatureInPreviewDistance - curvatureNextPosition	Data about the lane where the middle of the front of the agent is (e.g. mainLane) - Wether there is a lane on this position - Curvature at current s position - Width at current s position - Distance from current position to the end of the lane - Curvature at previewDistance - Curvature at position the agent will be in the next time step (it constant velocity)
laneRight - exist - curvature - width - distanceToEndOfLane - curvatureInPreviewDistance - curvatureNextPosition	Data about the lane to the right of the mainLane (in driving direction) - Wether there is a lane on this position - Curvature at current s position - Width at current s position - Distance from current position to the end of the lane - Curvature at previewDistance - Curvature at position the agent will be in the next time step (it constant velocity)
laneRightRight - exist - curvature - width - distanceToEndOfLane - curvatureInPreviewDistance - curvatureNextPosition	Data about the lane two lanes right of the mainLane (in driving direction) - Wether there is a lane on this position - Curvature at current s position - Width at current s position - Distance from current position to the end of the lane - Curvature at previewDistance - Curvature at position the agent will be in the next time step (it constant velocity)

Features about lane safety and lane change safety

Features about lane safety

Feature	Meaning
IsLaneSafe	A lane is safe to change to
IsSideLaneSafe	A lane is safe to change into from a normal situation
IsCurrentLaneSafe	The distance to the front and rear vehicle exceedes the minimum time headway

Features about lane change safety

Feature	Meaning
IsLaneChangeSafe	The current lane change is still save to perform
IsLaneSafeBeingQueued	A lane is safe to change to within traffic jams

ScmDependenciesInterface

Provides methods to update dependencies for SCM and grants access to them.

ScmDependencies

DriverParameters
VehicleModelParameters
StochasticsInterface
SpawnVelocity
CycleTime
OwnVehicleInformationSCM
SurroundingObjectsSCM
MesoscopicInformationSCM
TrafficRuleInformationSCM
GeometryInformationSCM

ScmComponents

ScmComponents is a container for components or modules SCM consists of (MentalModel, SituationManager, etc..) and is reponsible to provide necessary dependencies for these. For example, the MentalModel needs all ScmDependencies but the GazeFollower only needs OwnVehicleInformationScm.

MentalModel

The MentalModel presents a mental representation of the environment that the driver builds up through sensory perception. It receives the perceived information from the submodules InformationAcquisition - own vehicle information and Auditory Perception. The contents of the mental model can be queried with the help of different submoduls from the cognitive submodules Detailed descriptions of situation clusters, ActionManager and ActionImplementation. In the case of missing information or insufficient information quality, extra information requests can also be delivered into the mental model,which allow a targeted influencing of visual attention in the submodule Gaze Control.

FeatureExtrator

FeatureExtractor extracts features in the vicinity of the driver. The FeatureExtractor holds static functionality to extract salient features of the situation that the driver in question is surrounded by. Most methods extract these features from the mental model of the driver.