removed redundant osc2 files

README.md

@@ -46,6 +46,7 @@ Running CAWSR
 CAWSR can both be ran `locally` or `distributed`. Due to the high-spec requirements, it is recommended to run distributed if you do not meet the following minimum specs:
 - At least **10GB** VRAM and a modern GPU (2080 ti or newer)
     - We currently supports 20 and 30 series GPUs. Compatibility with the 40 series is current WIP.     
+
 - At least **32GB** RAM
 - A modern Intel or AMD CPU with at least 8 cores.
 

cawsr.py

@@ -37,6 +37,7 @@
 from srunner.scenarioconfigs.route_scenario_configuration import (
     RouteScenarioConfiguration,
 )
+from srunner.tools.route_visualisation import visualise_route
 from srunner.tools import route_manipulation
 from srunner.objects.ego_vehicle import EgoVehicle
 from srunner.tools.log import LogUtil
@@ -161,11 +162,24 @@ def run_scenario(
         # update the world
         CarlaDataProvider.set_world(self.carla_world)
 
-        logger.info("Spawning ego...")
-        ego = EgoVehicle(env_config)
-        actor = ego.spawn()
-        self.ego_vehicles.append(actor)
-        logger.info(f"Spawned ego with id: {actor.id}")
+        # attempt to spawn the ego 3 times for redundancy
+        max_attempts = 3
+        for attempt in range(max_attempts):
+            try:
+                logger.info("Spawning ego...")
+                ego = EgoVehicle(env_config)
+                actor = ego.spawn()
+                self.ego_vehicles.append(actor)
+                logger.info(f"Spawned ego with id: {actor.id}")
+                break  # exit loop if successful
+            except Exception as e:
+                logger.error(f"Failed to spawn ego on attempt {attempt + 1}: {e}")
+                if attempt == max_attempts - 1:
+                    logger.error(
+                        "Failed to spawn ego: Terminating scenario. This is most likely an issue with CARLA"
+                    )
+                    raise
+                time.sleep(2)
 
         # client must tick to spawn actors
         self._tick_carla()
@@ -194,6 +208,9 @@ def run_scenario(
         )
         route_config.agent.set_global_plan(gps_route, route)  # set agent route
 
+        # visualise the route in CARLA
+        visualise_route([waypoint[0] for waypoint in route])
+
         ego.prepare_ego(route[0][0])  # set location to first waypoint
 
         # allow the agent X ticks to initialize sensors and set the route
@@ -420,6 +437,7 @@ def _cawsr_process(
         else:
             seed = 0
 
+        env_config.iteration = run
         scenario_process = multiprocessing.Process(
             target=self.run_scenario,  # need to catch connection exception
             args=(
@@ -446,11 +464,12 @@ def _cawsr_process(
         self.results_manager.cleanup_xml()
 
         # copy over the recording from CARLA container
-        time.sleep(1)  # ensure file is written
+        time.sleep(1)  # ensure file is finished writing
         CARLAManager.fetch_file(
             "/home/carla/recording.log",
             self.results_manager.last_scenario,
         )
+        time.sleep(1)  # allow some time for docker to copy it over
 
         # fetch execution status of the scenario (failure or success)
         status = result["status"]

paper.md

@@ -34,7 +34,7 @@ bibliography: paper.bib
 
 # Summary
 
-`CAWSR` (**C**ARLA-**A**uto**W**are-**S**cenario **R**unner) facilitates the simulation-based testing of the open-source autonomous driving system, Autoware, within CARLA, the state-of-the-art open-source driving simulator. Building on existing tools, this project introduces a research-oriented testing framework for the execution of complex driving scenarios, as well as supporting implementation of a wide range of verification strategies.
+`CAWSR` (**C**ARLA-**A**uto**W**are-**S**cenario **R**unner) facilitates the simulation-based testing of the open-source autonomous driving system, Autoware, within CARLA, the state-of-the-art open-source driving simulator. Building on existing tools, this project introduces a research-oriented testing framework for the execution of complex driving scenarios, as well as supporting the implementation of a wide range of verification strategies.
 
 # Statement of Need
 
@@ -67,10 +67,23 @@ This facilitates a wide range of verification strategies based on common metrics
 Lastly, it is worth noting that simulators can often introduce unintended nondeterminism, which leads to inconsistent test results [@9793395; @osikowicz2025empirically].
 Therefore, `CAWSR` is designed to minimise such nondeterminism throughout the evaluation pipeline.
 
+# Research Impact
 
-# Tool Overview
+`CAWSR` provides a significant research impact by bridging the gap between the widely-used CARLA simulator and the industry-grade Autoware ADS.
+It addresses a critical bottleneck in the ADS testing research community by offering a modern replacement for the outdated Apollo/LGSVL workflow, which has lacked official support since 2022.
 
-`CAWSR` is a fully synchronous testing framework that directly integrates the CARLA simulator, Scenario Runner (as the scenario executor), and Autoware (as the System Under Test) to facilitate autonomous driving testing research. The tool is distributed as a containerized deployment using Docker and currently supports two modes of operation:
+It enhances research reproducibility by implementing a fully synchronous evaluation pipeline that minimises unintended nondeterminism in simulation-based testing.
+To ensure community readiness and ease of use, it is distributed as a Docker container.
+
+Furthermore, by adopting CARLA Leaderboard metrics, `CAWSR` enables researchers to directly compare Autoware with other state-of-the-art driving agents in the CARLA environment.
+It provides an essential foundation for the systematic evaluation of various testing approaches for ADS.
+
+
+
+# Software Design
+<!-- I've renamed the "Tool Overview" section since it fits quite well here, and added a few justifications.-->
+
+`CAWSR` is a fully synchronous testing framework that directly integrates the CARLA simulator, Scenario Runner (as the scenario executor), and Autoware (as the System Under Test) to facilitate autonomous driving testing research. The tool is distributed as a containerized deployment using Docker to manage complex dependencies and simplify the setup process. Currently, two modes of operation are supported:
 
 1. *Scenario Generation Mode:* Enables the dynamic generation and execution of scenarios (e.g. iterative scenario generation) provided by a user-defined algorithm. This is particularly useful for assessing the performance of new simulation-based ADS testing techniques.
 2. *Benchmark Mode:* Allows the execution of a predefined set of scenario definitions provided by the user. This is useful for standardised evaluations and comparisons between different driving agents.
@@ -85,7 +98,8 @@ The evaluation pipeline is engineered to be fully synchronous, minimising uninte
 
 - JSON Parser: Translates the *scenario_definition* (see \autoref{fig:scenario_domain}) into a Behavior Tree (BT). It utilises Scenario Runner's *Atomic Behaviours* and *Atomic Conditions* as modular primitives to define discrete actions (e.g., spawning pedestrians) and logic triggers.
 
-- ScenarioManager: Orchestrates the simulation loop by evaluating the BT to update actor states and triggering CARLA simulation ticks. Execution terminates based on CARLA Leaderboard criteria [@carla_leaderboard], as summarised in \autoref{tab:termination_criteria}. Post-execution, the module calculates the Driving Score (DS) according to the official leaderboard metrics.
+<!-- David: Citing SR for added transparency, since this class is natively part of it -->
+- ScenarioManager [@carla_scenario_runner_2025]: Orchestrates the simulation loop by evaluating the BT to update actor states and triggering CARLA simulation ticks. Execution terminates based on CARLA Leaderboard criteria [@carla_leaderboard], as summarised in \autoref{tab:termination_criteria}. Post-execution, the module calculates the Driving Score (DS) according to the official leaderboard metrics.
 
 - Agent and CarlaBridge: The Agent manages the ROS2 connection to Autoware. At each timestep, the CarlaBridge [@carlaautowarebridge] transforms CARLA snapshots and sensor data into the Autoware coordinate system. Autoware processes these inputs to issue control commands, which the Agent then applies to the ego vehicle.
 
@@ -106,6 +120,10 @@ This model is based on the format introduced by Scenario Runner, facilitating su
 To summarise, `CAWSR` provides ADS testing research community an easy to use Autoware evaluation pipeline.
 We hope that this work can facilitate the evaluation of new testing approaches on a state of the art driving system.
 
+# AI Usage Disclosure
+
+<!-- David: I used AI to help explain concepts some concepts and generate short code snippets to get me started (mostly about ROS2 and networking). Is this the correct way to phrase it? -->
+Generative AI tools were used in this work solely to support high-level research concepts and structural ideas. All software implementation, including the source code, architecture, and deployment scripts, was authored entirely by the researchers without AI assistance.
 
 # Acknowledgements
 

srunner/autoagents/autoware_agent.py

@@ -17,6 +17,7 @@
 from srunner.autoagents.autoware_carla_interface.carla_autoware import (
     InitializeInterface,
 )
+from srunner.tools.CARLA_manager import CARLAManager
 
 import threading
 import rclpy
@@ -46,6 +47,7 @@ def setup(self, config: EnvironmentConfig) -> None:
         rclpy.init(args=None)
 
         self.config = config
+        self.tick_delta = CARLAManager.FIXED_DELTA_SECONDS
 
         # initialise autoware state object
         self.autoware_state = autoware_state.AutowareState("ego_vehicle", None)
@@ -56,7 +58,9 @@ def setup(self, config: EnvironmentConfig) -> None:
         self.carla_interface = InitializeInterface(self.config, self._node)
 
         self.state_node = state_node.StateNode(self.autoware_state, self._node_state)
-        self.state_node.reset_autoware(self.config.town, self.config.ego_name)
+        if config.iteration == 1:
+            # only reset autoware on the first run
+            self.state_node.reset_autoware(self.config.town, self.config.ego_name)
 
         self.route_node = route_node.RouteNode(self.autoware_state, self._node_state)
         self.autoware_node = autoware_node.AutowareNode(
@@ -92,7 +96,6 @@ def set_route(self) -> None:
         self.goal_pose_world = self._global_plan_world_coord[-1]
         self.waypoints_world = self._global_plan_world_coord[:-1]
 
-        # autoware cannot handle many waypoints, becomes unreliable
         n_waypoints = len(self.waypoints_world)
         segment_size = int(n_waypoints / 3)
 
@@ -172,7 +175,7 @@ def run_step_init(self) -> bool:
     def run_step(self) -> None:
         """Tick method containing all logic based on autoware state"""
         self.counter += 1
-        if self.counter % 20 == 0:
+        if self.counter % int(1 / self.tick_delta) == 0:
             logger.info(
                 f"Ticked 1 second game-time, actual tick is {(time.perf_counter_ns() - self.last_tick) / 1e6}ms"
             )

srunner/autoagents/autoware_carla_interface/carla_ros.py

@@ -26,7 +26,6 @@
 from geometry_msgs.msg import Pose
 from geometry_msgs.msg import PoseWithCovarianceStamped
 import numpy
-import datetime
 import pathlib
 from rosgraph_msgs.msg import Clock
 from sensor_msgs.msg import CameraInfo
@@ -54,6 +53,8 @@
     SensorInterface,
 )
 
+from srunner.tools.CARLA_manager import CARLAManager
+
 
 class carla_ros2_interface(object):
     def __init__(self, node):
@@ -80,9 +81,11 @@ def __init__(self, node):
             "pose": 2,
         }
         self.publish_prev_times = {
-            sensor: datetime.datetime.now() for sensor in self.sensor_frequencies
+            sensor: GameTime.get_time() for sensor in self.sensor_frequencies
         }
 
+        self.delta = CARLAManager.FIXED_DELTA_SECONDS  # delta in seconds
+
         self.ros2_node = node
 
         self.clock_publisher = self.ros2_node.create_publisher(Clock, "/clock", 10)
@@ -137,6 +140,11 @@ def __init__(self, node):
                 self.pub_camera_info = self.ros2_node.create_publisher(
                     CameraInfo, "/sensing/camera/traffic_light/camera_info", 1
                 )
+
+                # self.pub_camera_yolo_info = self.ros2_node.create_publisher(
+                #   CameraInfo, "/sensing/camera/0/camera_info", 1
+                # )
+
             elif sensor["type"] == "sensor.lidar.ray_cast":
                 if sensor["id"] in self.sensor_frequencies:
                     self.pub_lidar[sensor["id"]] = self.ros2_node.create_publisher(
@@ -165,9 +173,9 @@ def __call__(self):
         return control
 
     def checkFrequency(self, sensor):
-        time_delta = (
-            datetime.datetime.now() - self.publish_prev_times[sensor]
-        ).microseconds / 1000000.0
+        # implement frequency check based on game time
+        time_delta = GameTime.get_time() - self.publish_prev_times[sensor]
+
         if 1.0 / time_delta >= self.sensor_frequencies[sensor]:
             return True
         return False
@@ -185,7 +193,7 @@ def lidar(self, carla_lidar_measurement, id_):
         """Transform the received lidar measurement into a ROS point cloud message."""
         if self.checkFrequency(id_):
             return
-        self.publish_prev_times[id_] = datetime.datetime.now()
+        self.publish_prev_times[id_] = GameTime.get_time()
 
         header = self.get_msg_header(frame_id="velodyne_top_changed")
         fields = [
@@ -258,7 +266,7 @@ def pose(self):
         """Transform odometry data to Pose and publish Pose with Covariance message."""
         if self.checkFrequency("pose"):
             return
-        self.publish_prev_times["pose"] = datetime.datetime.now()
+        self.publish_prev_times["pose"] = GameTime.get_time()
 
         header = self.get_msg_header(frame_id="map")
         out_pose_with_cov = PoseWithCovarianceStamped()
@@ -335,7 +343,7 @@ def camera(self, carla_camera_data):
 
         if self.checkFrequency("camera"):
             return
-        self.publish_prev_times["camera"] = datetime.datetime.now()
+        self.publish_prev_times["camera"] = GameTime.get_time()
 
         image_data_array = numpy.ndarray(
             shape=(carla_camera_data.height, carla_camera_data.width, 4),
@@ -356,7 +364,7 @@ def imu(self, carla_imu_measurement):
         """Transform a received imu measurement into a ROS Imu message and publish Imu message."""
         if self.checkFrequency("imu"):
             return
-        self.publish_prev_times["imu"] = datetime.datetime.now()
+        self.publish_prev_times["imu"] = GameTime.get_time()
 
         imu_msg = Imu()
         imu_msg.header = self.get_msg_header(frame_id="tamagawa/imu_link_changed")
@@ -416,7 +424,7 @@ def ego_status(self):
         if self.checkFrequency("status"):
             return
 
-        self.publish_prev_times["status"] = datetime.datetime.now()
+        self.publish_prev_times["status"] = GameTime.get_time()
 
         # convert velocity from cartesian to ego frame
         trans_mat = numpy.array(self.ego_actor.get_transform().get_matrix()).reshape(

srunner/autoagents/autoware_carla_interface/modules/carla_wrapper.py

@@ -23,6 +23,8 @@
 
 from srunner.scenariomanager.carla_data_provider import CarlaDataProvider
 
+logger = logging.getLogger("scenario-runner")
+
 
 # Sensor Wrapper for Agent
 class SensorReceivedNoData(Exception):
@@ -54,7 +56,7 @@ def __call__(self, data):
         elif isinstance(data, GenericMeasurement):
             self._parse_pseudo_sensor(data, self._tag)
         else:
-            logging.error("No callback method for this sensor.")
+            logger.error("No callback method for this sensor.")
 
     # Parsing CARLA physical Sensors
     def _parse_image_cb(self, image, tag):
@@ -216,6 +218,13 @@ def setup_sensors(self, vehicle, debug_mode=False):
                     yaw=sensor_spec["spawn_point"]["yaw"] + 0.0364,
                 )
 
+            bp.set_attribute(
+                "sensor_tick", str(round(1.0 / sensor_spec["publish_frequency"], 2))
+            )
+            logger.info(
+                f"Sensor {sensor_spec['id']} tick rate set to {str(round(1.0 / sensor_spec['publish_frequency'], 2))}"
+            )
+
             # create sensor
             sensor_transform = carla.Transform(sensor_location, sensor_rotation)
             sensor = CarlaDataProvider.get_world().spawn_actor(

srunner/autoagents/autoware_carla_interface/objects/sensors.json

@@ -13,7 +13,8 @@
       },
       "image_size_x": 1920,
       "image_size_y": 1080,
-      "fov": 90.0
+      "fov": 90.0,
+      "publish_frequency": 11.0
     },
     {
       "type": "sensor.lidar.ray_cast",
@@ -32,7 +33,8 @@
       "upper_fov": 10.0,
       "lower_fov": -30.0,
       "rotation_frequency": 20,
-      "noise_stddev": 0.0
+      "noise_stddev": 0.0,
+      "publish_frequency": 11.0
     },
     {
       "type": "sensor.other.gnss",
@@ -44,7 +46,8 @@
         "roll": 0.0,
         "pitch": 0.0,
         "yaw": 0.0
-      }
+      },
+      "publish_frequency": 2.0
     },
     {
       "type": "sensor.other.imu",
@@ -56,7 +59,8 @@
         "roll": 0.0,
         "pitch": 0.0,
         "yaw": 0.0
-      }
+      },
+      "publish_frequency": 50.0
     }
   ]
 }

srunner/scenarioconfigs/environment_configuration.py

@@ -26,3 +26,4 @@ def __init__(self) -> None:
         self.ego_spawn: carla.Transform | None = None
         self.sensor_config: list[DefaultSensor] = []
         self.route_id: int = 0
+        self.iteration: int = 0