Skip to content

Road Side Units (RSU) in Simple-AV

In the Simple-AV project, Road Side Units (RSUs) play a crucial role in the V2X (Vehicle-to-Everything) communication setup. These units are strategically placed at intersections to enhance the detection coverage and reduce blind spots, creating a smart infrastructure within the simulation environment.

RSUs as Pseudo Sensors

One of the key innovations we implemented in our V2X_E2E project is the integration of RSUs as Pseudo Sensors. Unlike traditional sensors that use point cloud data (PCD) and perform heavy object detection pipelines, pseudo sensors aim to significantly reduce the computational load while maintaining detection accuracy.

How Pseudo Sensors Work

Instead of using complex object detection algorithms or raw PCD from LiDARs, pseudo sensors operate based on assigned points on objects.

  • When we create objects like vehicles or pedestrians in the simulation, we assign a specific number of detection points to their physical model.
  • The number of points is proportional to the object's size. For instance:
  • A truck may have around 30 points.
  • A pedestrian may have 10 points.

The pseudo sensor then radiates rays in the environment, and if the rays hit enough points on an object (exceeding a predefined detection threshold), the sensor considers the object detected.

For example: - If a truck passes through the intersection and the sensor detects at least 10 points on it, the object is marked as detected. - Once detected, the sensor publishes detailed information about the object, including: - Position - Rotation - Linear Speed - Type (e.g., pedestrian, car, truck)

Advantages of Pseudo Sensors

  • Low Computation Cost: Avoids the need for real-time LiDAR point cloud processing or complex object detection algorithms.
  • Efficient Detection: Suitable for intersections where real-time data from multiple objects is crucial.
  • Smart Infrastructure: Enables the simulation of smart intersections with enhanced V2X capabilities.

Creating a Smart Intersection

We use pseudo sensors to build a setup of RSUs at each intersection, forming a smart infrastructure. This setup allows us to: - Test sensor placements to maximize intersection coverage. - Minimize blind spots by strategically positioning RSUs. - Study the impact of smart infrastructure on traffic safety and efficiency.

Example of Object with Detection Points

Here are some examples of objects with assigned detection points:

image

By leveraging RSUs as pseudo sensors, we create a more efficient and realistic V2X simulation that is computationally lightweight while still providing accurate detection at intersections.