Testing Racing-Stack with Simulation

This guide explains how to run Racing-Stack in a simulation environment, allowing you to test without an actual vehicle.

You can use this for logic verification and testing before operating the actual vehicle.

Mapping

1. Since you cannot create a map within the simulation environment, you need map.png and map.yaml files.

Note) File generation process on the actual vehicle:

1. Run roscore
2. Launch vehicle sensors (LIDAR, IMU, VESC) with roslaunch stack_master low_level.launch
3. Generate files with roslaunch stack_master mapping.launch map:=map_name create_map:=true (drive one lap around the closed-loop map)

File Example

Field	Description
image	Filename
resolution	1 pixel = 0.05m
origin	Map origin point
occupied_thresh	Obstacle detection threshold
free_thresh	Free space detection threshold
negate	Image inversion flag

1. Calculate the optimal path based on the map.

roslaunch stack_master mapping.launch map:=map_name create_map:=false create_global_path:=true

• Extract centerline based on map.png

• Set Speed sector for speed tuning per section

• Set Overtaking sector to define overtaking allowed zones

Three files are generated in the end.

• global_waypoints.json: Optimal path and speed profile
• speed_scaling.yaml: Speed limits per section
• ot_sectors.yaml: Overtaking allowed zones

Head to head

1. Launch the simulation environment.

roslaunch stack_master base_system.launch map:=map_name sim:=true

1. Enable obstacle detection, driving state judgment, and vehicle control functions.

roslaunch stack_master headtohead.launch perception:=false

Obstacle

1. Static obstacle avoidance

You can create static obstacles using Publish Point in rviz.

The system detects the static obstacle and generates an avoidance path, showing collision-free obstacle avoidance.

2. Dynamic obstacle overtaking

You can create a dynamic obstacle that travels along the optimal path at a slower speed (0.4x).

roslaunch obstacle_publisher obstacle_publisher.launch speed_scaler:=0.4

The system detects the dynamic obstacle and generates an overtaking path, demonstrating successful overtaking behavior.

Conclusion

This guide covered how to test Racing-Stack in a simulation environment.

• You can verify racing-stack logic using simulation.
• You can generate map-based driving data (optimal path, speed profile, overtaking zones).
• You can test virtual obstacle scenarios (static obstacle avoidance and dynamic obstacle overtaking).

We recommend thoroughly verifying in simulation before deploying to the actual vehicle.