- configurable submodules allowing open or hierarchical module composition
- Modularized adore scheduler@ github.com
- modularized adore cli@ github.com/DLR-TS/adore_cli.git
- modularized libadore@ github.com/DLR-TS/libadore
- modularized adore_if_ros@ github.com/DLR-TS/adore_if_ros
- updated adore_if_carla submodule to tag v0.1.0
- bugfixes to the build system
- introduced caching of externally provided docker images in folder .docker_image_cache
- modularized respository structure
- added osqp, lizard, cppcheck, cpplint
- added adore_if_carla
- docker wrappers and make files
- externalized pngs
- New node adore_crosstraffic_provider: Monitors prediction of other traffic participants and tests for intersection of cross-traffic with lane following view. In case of crosstraffic intersecting the lane following view, the nearest intersection with a feasible stopping position is published on the topic ENV/ConflictSet. The node may be extended in the future to publish also all intersection points after the first potential stopping position.
- adore_trajectory_planner_lf_node has been extended with several startup parameters:
speed_scale \in [0.0,1.0]sets the reference speed scale for the planner. With 1.0 the planner tries to achieve the speed limit, with 0.5 the planner tries to achieve 50% of the speed limit, with 0.0 the planner tries to stop the vehicle.lateral_i_grid \in [...,-1,0,+1,...]sets the reference offset from the lane center.stop_point_id \in [-1,0,1,...]sets the index of the stopping position from ENV/ConflictSet, which the planner tries to achive. Currently-1and0are selectable, where-1indicates all stopping positions to be ignored and0forces the planner to stop at the first stopping position.
- added notion of static obstacles, represented as OccupancyCylinder msgs, enables seperate notification for traffic based obstacles and static obstacles
- Adding functionality for separating odometry and localization information, so that feedback-control works on odometry information and feed-forward (trajectory planning) works on localization information. adore_vehiclemodel_node is adapted to output ground-truth and optionally localization and odometry information for error free simulation. If simulating errors, localizationmodel and odometrymodel consume ground-truth and output signals superimposed with typical errors on localization and odometry topics. Feed-back control is modified to operate on odometry information. Tactical planners are modified to monitor transformations between localization and odometry. Tactical planners select initial state for planning request in such a way that initial state is jump free in odometry coordinate system. Trajectory planners do not have to be modified, as operating on modified input from tactical planners is sufficient to guarantee consistency of initial state. Tactical planners provide the executed trajectory in odometry and localization coordinates.
- adore_if_ros: src/adore_localizationmodel_node.cpp - simulates localization errors: error random walk and localization jumps (model for gps multipath effect)
- adore_if_ros: src/adore_odometrymodel_node.cpp - simulates odometry errors: initialized at (0,0), drift error
- libadore: fun/setpointrequest_dispatcher.h - selects initial state for trajectory planning, assures consistency of initial state and provides setpointrequest in localization as well as odometry coordinates.
- adore_if_ros_demos/demo013_localization_errors.launch - demonstrates improved behavior under occurance of localization jumps
- Added speedlimit provider node and notion of speedlimit functions to the laneview_provider
- adore_if_ros: conversions/cooperativeuserspredictionconverter.h
- adore_if_ros: conversions/platooninginformationconverter.h
- adore_if_ros: adore_pvprovider_node.cpp
- adore_if_v2x: mcm_to_platoon_node.cpp
- adore_if_ros_msg: CooperativePlanning.msg
- adore_if_ros_msg: CooperativePlanningSet.msg
- adore_if_ros_msg: CooperativeUser.msg
- libadore: apps/platoon_view_provider.h
- libadore: env/traffic/cooperativeusersprediction.h
- libadore: env/traffic/cooperativeuserprocess.h
- libadore: fun/ctrl/platoon_controller.h
- libadore: fun/logic/distanceStateMachine.h
- libadore: fun/logic/platoonLogic.h
- libadore: fun/logic/platoonStateMachine.h
- libadore: fun/platooningInformation.h
mcm_dmove is used for adore_if_v2x:mcm_to_prediction_node and adore_if_v2x:setpoint_to_mcm mcm_dmove is used instead of SimMCM in adore_if_v2x
Predictions could contain nan values and would thus lead to false positive collision detections. Error was caused by errors in map: If lanes exist with only one point per border, the computed centerlane has no points at all. Prediction was changed to test for nan and discard if nan occurs.
- libadore: env/traffic/gapdata.h
- libadore: env/traffic/lanechangegaps.h
- libadore: apps/gapprovider.h
- adore_if_ros_msg: GapData.msg
- adore_if_ros: adore_gap_provider.cpp
adore_trajectory_planner_lc_node uses basiclanechangeplanner.h, which restricts lane changes in such a way that front vehicles may not be overtaken. Passing the (former) front vehicle would occur during the succeeding lane-following maneuver. A new planner adore_trajectory_planner_alc_node with advancedlanechangeplanner.h has been added: The new planner allows for overtaking of the front vehicle during the lane change. On the one hand the advanced version is more dynamic and improves assessment wrt. progress in the tactical planner. On the other hand the criticality of advanced overtaking maneuver is higher. Both planners should be maintained for the future, in order to meet specific project requirements.
-
Migration to Ubuntu 20.04 and ROS noetic
-
New support for Road2Simulation map format
-
Several new modules have been added with this version, which are discussed in the following:
-
adore_if_ros_demos: New application examples
- demo007a_precedence_right.launch demonstrates defining and adhering to right of way with the help of road-based predictions
- demo007b_precedence_no_rules.launch serves as comparison to 007a: Behavior with undefined right of way
- demo007c_precedence_left.launch demonstrates definition of right of way opposite to case 007a
- demo007d_sl45_right_turn_pedestrian.launch demonstrates non-road-based prediction of traffic participant and conflict resolution
- demo008_sumo_trafficlights.launch demonstrates definition of scenarios with traffic lights controlled by sumo and simulation of MAPEM and SPATEM V2X messages, as well as reaction of vehicle to messages
- demo009_lanechange_tostmannplatz.launch demonstrates lane change maneuver planning and maneuver selection by tactical planner on a realistic urban roadway
- demo010_sat_images.launch demonstrates how to integrate satellite image data from wms geo service for visualization
- demo011_scheduler.launch demonstrates faster/slower than real-time as well as headless simulation for "batch processing" of scenarios
-
adore_if_ros: New vehicle control nodes
- adore_checkpoint_controller_node - ROS node, which monitors a predefined list of checkpoints and enforces driver interaction before vehicle may continue over a given checkpoint
- adore_lvprovider_node - ROS node, which computes adore::views (lane following and lane change) and provides these to other adore processes
- adore_mission_controller_node - ROS node with vehicle's top level state automaton responsible for switching between driving/parking and next goal location
- adore_monitor0_node - ROS node which computes some zero order logic propositions about vehicle state, e.g. IN_COLLISION
- adore_prediction_filter - ROS node which filters predictions for relevance to the ego vehicle
- adore_prediction_provider - ROS node which comutes motion predictions in form of occupied space time for traffic participants detected near ego vehicle
- adore_tactical_planner_node - ROS node which requests plans from trajectory planners and selects one of the results for execution
- adore_trajectory_planner_lc_node - ROS node which plans lane change trajectories
- adore_trajectory_planner_lf_node - ROS node which plans lane following trajectories
- adore_trajectory_planner_lm_node - ROS node which plans lane merge trajectories (finalization of lane change maneuver)
- area_of_effect_provider - ROS node which computes area that is accessible to ego vehicle
- plainxmlexporter - An offline, std cpp, standalone process, which converts from ADORe map formats to SUMO plainXML transfer format
-
adore_if_ros: Changes to plotting tools (for plotlab)
- adore_borderbird_node - deleted
- plot_graph_node - ROS node for graph plotting of numeric ROS topics with plotlab
- plot_area_off_effect - visualize output of area_of_effect_provider
- plot_ego_node - visualize ego vehicle as simple line drawing or .png image
- plot_lanes_node - visualize lanes near ego vehicle
- plot_plan_swath_node - visualize space covered by ego when following planned trajectories
- plot_predictions_node - visualize prediction of space covered by detected traffic participants
- plot_satimages_node - displays sattelite images in background of top view plot; images are downloaded from specified geoserver
- plot_traffic_node - visualize detected traffic participants as simple line drawings or .png images
- plot_trajectory_node - visualize all currently available trajectories
- plot_views_node - visualize lane-following and lane-change views
-
adore_if_ros: Some auxiliary nodes for testing
- test_control_dashboard - plots some values, which are insightful for debugging and controller tuning
- test_control_error_node - allows to induce control errors in order to test driver reaction times
- test_lc_trajectory_planner_node
- test_mrm_planner_node
- test_setspeedlimit - allows to insert speed-limit changes at certain locations
- test_straight_line_prediction_node
- test_trajectory_planner_node
- test_vectorsize.py - rostopic style command line tool: displays size of vectors contained in ROS messages
- plotlab2navigationgoal.py - publishes navigation goal, if a coordinate is selected in plotlab (pressing 'l')
-
adore_if_ros_msg (ROS messages for ADORe ROS interface)
- AreaOfEffect.msg - determines area that could be affected by vehicle, allows to filter objects accordingly
- CostPair.msg - pair of objective value name and objective value, used in PlanningResult.msg
- LaneChangeGeometry.msg - encodes information for adore::view::ALaneChangeView
- LaneFollowingGeometry.msg - encodes information for adore::view::ALane
- LaneGeometry.msg - combines LaneChangeGeometry.msg and LaneFollowingGeometry.msg
- LinearPiecewiseFunction*d.msg - linear, piecewise functions
- MissionState.msg - state information for mission controller
- OccupancyCylinder.msg - Space time volume for collision detection system
- OccupancyCylinderPrediction.msg - A predicted sequence of OccupancyCylinders, can be associated with a tracked object and can form the branch of a prediction tree
- OccupancyCylinderPredictionSet.msg - A set of OccupancyCylinderPredictions
- PlanningRequest.msg - Used to request trajectory planners to start planning for a given initial state
- PlanningResult.msg - A trajectory planner's result, reply to PlanningRequest.msg
- PlanningResultSet.msg - A set of PlanningResult.msg
- PlotStart.msg - A message notifying plotter to start plotting numeric values of given ROS topics
- TCDConnectionStateTrace.msg - The signal state over time for a singel connection path at an intersection
-
v2x_if_ros (V2X ROS interface)
- autogenerated ROS messages for ETSI ITSG5 standardized and test V2X radio messages
- CAM v2 (standardized) - publish vehicle state
- CPM v1_19 (standardized) - distributed awareness, publish vehicle observations
- DSRC v2 (standardized) (MAPEM, SPATEM, SREM, SSEM) - intersection topology and traffic light signal phase timing, signal phase requests
- MCM transaid - maneuver coordination message, publish vehicle intended behavior, request vehicle behavior adaptation
- MCM dmove - extension of MCM transaid version: request vehicle to follow specific trajectory
- ODDsinfo - publish operational domain information
- STRP fau - space time reservation for vehicle motion coordination
- v2x_sim: simulation of v2x radio message transmission
- channel_sim_node - ROS node for handover between a simulated station and the simulated V2X radio channel
-
adore_if_v2x (ADORe interface to ROS V2X messages)
- mcm_to_prediction_node - ROS node for conversion of a received MCM V2X message into a vehicle prediction processable by ADORe vehicle
- setpointrequest_to_mcm_node - ROS node for publishing the trajectory that will be executed by ADORe vehicle
- v2x_trafficlights_node - ROS node for conversion of MAPEM and SPATEM V2X messages into signal state information for ADORe vehicle
-
sumo_if_ros (eclipse SUMO interface to ROS)
- sumotls2ros_node - ROS node, which extracts traffic light signal phases from SUMO via TraCI and publishes the information as MAPEM and SPATEM messages
-
keepmoving_if_ros: Communication with REST server to aquire mission goals and provide information about ego vehicle. (preliminary devolpment stage, future changes are expected)
- shuttle.py - monitors ego vehicle state and pushes info to REST server for fleet management
- journey.py - requests state, route and target information from REST server
- Updated lane view provider node, does now support lane following and lane change views
- New machine learning functionality allows training of reinforcement learning agents.
- Modifications of IP check
- removed license information from dependencies.md files
- removed comments from package.xml files and fixed license string
- removed stb_image.h from plotlabserver repository and created automatic download triggered by cmake
- Initial Eclipse release
- New ROS nodes
- adore_if_ros/adore_borderbird_node - simple visualization a simulated vehicle's local view
- adore_if_ros/adore_feedbackcontroller_node - stabilization of vehicle around reference trajectories: linear feedback controller
- adore_if_ros/adore_lfbehavior_node - lane-following-only planner (no lane changes)
- adore_if_ros/adore_objectdetectionmodel_node - a simple (exemplary) sensor model for detection of other traffic participants in simulation
- adore_if_ros/adore_timer_node - time signal for simulation, pause simulation
- adore_if_ros/adore_vehiclemodel_node - simulation of a simple vehicle model (linear bicycle model)
- sumo_if_ros/sumotraffic2ros_node - a node which synchronizes simulation in SUMO and ADORe vehicles in ROS
- New demos
- demo003: Simulation of an ADORe automated vehicle, which follows the lanes of an example map
- demo004: Simulation of three ADORe automated vehicles on a circular track
- demo005: Simulation of an ADORe automated vehicle and several SUMO vehicles on a circular track
- demo006: Simulation of an ADORe automated vehicle, which navigates to a goal position by lane following and turning at junctions
- Initial Eclipse release
- libadore/adore/sim: Simulation tools
- libadore/adore/fun: Automated driving control functions, qpOASES based maneuver planner, feed-back control
- libadore/adore/apps, new applications:
- borderbird.h - simple birds-eye-view plotting application connecting to plotlabserver
- feedbackcontroller.h - trajectory tracking
- lane_following_behavior.h - application of qpOASES based maneuver planner to lane following
- monitor0.h - prototype application for evaluation of important predicates for automatic simulation evaluation
- objectdetectionmodel.h - application with simple object detection model
- plainxmlexporter.h - prototype application for exporting of maps to sumo plain xml format
- test_lc_trajectory_planner.h - test application for lane change maneuver planning with lots of visualizations
- test_trajectory_planner.h - test application for lane following maneuver planning with lots of visualizations
- trafficlighsim.h - prototype application for stand-alone traffic light phase simulation
- vehiclemodel.h - application with vehicle model for use in simulation experiments
- Test release
- ROS interfaces for adore, ROS message definitions
- New ROS nodes
- adore_mapprovider_node - loading of OpenDrive tracks for automated vehicle
- adore_navigation_node - computation of a shortest route to a given target on loaded track
- New demos
- demo001: Loading of OpenDrive tracks
- demo002: Computation of distance functions on track
- libadore/adore/views: Data abstraction for nominal automated driving maneuver planning
- libadore/adore/env: BorderBased realization of LaneFollowingView and LaneChangeView
- libadore/adore/mad: Function library, nonlinear regression for fitting of road coordinate system, interface to qpOASES for dynamic optimization problems
- libadore/adore/if_xodr: OpenDrive file loader
- libadore/adore/apps, new applications:
- map_provider.h - streaming of static environment information in vehicle vicinity
- navigation.h - global navigation module / graph search