Testing

June 13th, 2025:
In this recent test, a follower Wave Adaptive Modular Vessel (WAM-V) successfully demonstrated real-time tracking and following of an uncooperative lead vessel using only onboard sensors and machine learning–based perception. The system operated without GPS, communication, or prior knowledge of the target, simulating conditions expected in adversarial or denied environments.

The follower USV used a combination of LiDAR point cloud clustering (K-Means) and a custom-trained YOLOv9 object detector to identify and track the lead vessel. A Kalman filter estimated motion in real time, and all decisions were processed onboard using ROS 2.

Key performance highlights from the test include:

• Reliable target detection at distances up to 25 meters

• Consistent 10-meter following distance maintained

• Zero use of GPS or inter-vehicle communication

• System operated fully onboard with passive sensing
  
  

This test marks a major milestone in demonstrating autonomy in operationally constrained maritime environments. It is part of Xavier Vicent’s dissertation research and supports development for future RoboBoat and RobotX competitions.

June 4th, 2025:
A simulation test was conducted using the Virtual RobotX (VRX) Gazebo environment to evaluate decentralized COLREGs-based navigation for multiple Unmanned Surface Vehicles (USVs). The objective was to validate real-time decision-making for encounter scenarios based on the International Regulations for Preventing Collisions at Sea (COLREGs), without relying on centralized planning or inter-vessel communication.

Each USV operated independently, using an onboard planner that combined Artificial Potential Fields (APF) with COLREGs logic to resolve dynamic interactions. Crossing, head-on, and overtaking scenarios were tested successfully, with each USV adapting its trajectory in real time based solely on onboard perception and local decision-making.

Key outcomes from the test include:

• Successful resolution of all encounter scenarios (head-on, crossing, overtaking)

• No inter-vessel communication or centralized control

• Smooth, COLREGs-compliant trajectories generated using APF-based shaping

• Real-time perception using simulated LiDAR and camera feeds in VRX

Note: This test did not yet include static obstacles such as buoys or docks. Integration of static object detection and classification into the COLREGs logic is planned for the next phase of development.

This work is part of Xavier Vicent’s dissertation research and supports future implementation in field-capable USV autonomy systems for RoboBoat and RobotX.

October 19th, 2024:

In this recent test, the Wave Adaptive Modular Vessel (WAMV) demonstrated highly accurate station keeping capabilities utilizing a PID controller with gains optimized through a Genetic Algorithm. The controller was tuned to ensure precise position holding in challenging environmental conditions, including wind gusts reaching up to 15 m/s.

Key performance metrics from the test include:

• Position accuracy: Less than 0.5 meters

• Heading accuracy: Less than 0.2 radians

• Test duration: 5 minutes of continuous station keeping

This test represents a significant milestone in our ongoing research efforts. The success of the PID station keeping controller is a crucial step toward achieving fully autonomous control for marine vehicles in complex environments.

This work is part of a research and thesis project conducted by team member Xavier Vicent, and it contributes directly to the development of advanced control strategies for the upcoming RobotX competition.

October 13th, 2024:
This is a successful first test flight of the Holybro X500 V2 drone. We conducted both manual flight and tested the drone's auto takeoff and auto land functionalities. The UAV performed flawlessly, responding well to manual controls and executing autonomous takeoff and landing as expected. This marks a major milestone in our testing process, as we now have a solid foundation for future autonomous operations.

The next step in our testing will focus on autonomous flight, where we will further refine the control systems and navigation algorithms to enable fully autonomous missions. The data gathered from today’s flight will be invaluable for optimizing the drone’s performance in future tests.

September 29th, 2024:

In our latest test on water, we are integrating bow thrusters into the RobotX WAM-V, enhancing its maneuverability and control. Alongside this addition, we are testing the bow thrusters in conjunction with the PID Station Keeping Controller.

This new configuration results in a more robust over actuated system, significantly improving performance during station keeping tasks. The bow thrusters provide enhanced lateral control, allowing for greater stability and precision in maintaining position in varying environmental conditions.

The data collected during this testing phase will be crucial for refining our control strategies and optimizing the overall performance of the WAM-V in upcoming challenges.

September 15th, 2024:

We are currently conducting two key tests, each demonstrated through a video, as part of our preparation for the RobotX 2024 Competition:

1. YOLO V9 Model for Buoy Detection:
   In the first video, we showcase our custom-trained YOLO V9 model detecting competition buoys using a stereo camera. This model demonstrates high accuracy in recognizing and localizing buoys, which is critical for successful navigation.

2. Point Cloud Detection and Path Simulation:
   The second video features a ROS bag playing a point cloud of competition buoys. We detect the buoys from the point cloud and generate a path between them, simulating the "Follow the Path" task for RobotX 2024. This approach allows for precise 3D mapping, improving obstacle avoidance and path planning.

These tests are pivotal in fine-tuning our systems for the competition, showcasing our advancements in autonomous detection and navigation.

September 3rd, 2024:

We are currently showcasing videos of our testing phase utilizing the VRX Simulator to evaluate the path planning capabilities for the tasks required in the upcoming RobotX 2024 Competition. These demonstrations feature custom worlds created within the Gazebo environment, simulating various maritime scenarios that our unmanned surface vehicle (USV) may encounter during the competition.

Through these video demonstrations, we aim to highlight the refinement of our path planning algorithms, ensuring that the WAM-V can navigate complex environments effectively and efficiently.

The data and insights gathered from these tests will be invaluable in preparing our systems for the challenges of the RobotX competition, where precision, adaptability, and autonomous decision-making are crucial for success.

July 9th, 2024:

In this test, our WAM-V USV demonstrated the seamless integration of an autonomous drone that took off, landed on a designated charging platform, recharged, and took off again—all autonomously. This evaluation is part of our ongoing research aimed at improving the coordination between unmanned surface vehicles and aerial drones, with a focus on energy management and precision landing systems. 

July 1st, 2024:

we are gathering data from both the LiDAR and camera systems for USV tracking of vehicles. This test aims to evaluate the effectiveness of combining sensor data to improve the accuracy and reliability of tracking moving objects in a maritime environment.

By analyzing the data collected from the LiDAR and camera, we aim to enhance our tracking algorithms, ensuring that our USV can effectively monitor and respond to dynamic situations in real time.

June 28th, 2024:

We are currently using the RobotX WAM-V to tune the PID Station Keeping Controller, a critical component for tasks in the upcoming RobotX competition. This controller is essential for launching and recovering a UAV, as well as performing the "Scan the Code" task.

In this test, the WAM-V demonstrates its ability to maintain stability and precision while navigating through various scenarios. By refining the PID parameters, we aim to enhance the vehicle's performance and reliability in autonomous operations.

The insights gained from this testing phase will play a significant role in preparing our systems for the challenges of the RobotX competition.

May 28th, 2024:

In this test, our WAM-V USV autonomously follows an L-shaped path using PID-based heading, speed, and station-keeping controllers. This evaluation is part of our ongoing research aimed at gathering crucial data for analyzing and refining PID-based control systems for USV navigation.

The data collected during this test will contribute to an upcoming publication focused on the performance and efficiency of PID controllers in maritime environments, enhancing the precision of autonomous vessel navigation and control.

April 5th, 2024:

In this test, our WAM-V USV autonomously follows an L-shaped path using PID-based heading, speed, and station-keeping controllers. This evaluation is part of our ongoing research aimed at gathering crucial data for analyzing and refining PID-based control systems for USV navigation.

The data collected during this test will contribute to an upcoming publication focused on the performance and efficiency of PID controllers in maritime environments, enhancing the precision of autonomous vessel navigation and control.

February 16th, 2024:

In this demonstration, our WAM-V USV showcases its advanced autonomous capabilities designed for sustainable maritime operations. The vehicle autonomously navigates to a designated location, deploys its anchor system for station-keeping, and then activates a water wheel to harness renewable energy from water flow. This energy is used to charge onboard batteries, enabling long-duration missions.

Once the battery charging is complete, the WAM-V retrieves its anchor and proceeds to the next set of waypoints, continuing its mission efficiently and sustainably.

This innovative system highlights the potential for autonomous maritime platforms in energy harvesting and environmental monitoring.

February 11th, 2024:

In this video, witness our WAM-V USV navigating the challenging channel task using LiDAR during the final run of the 2024 RoboBoat Competition. Utilizing LiDAR for precise environmental mapping, the USV autonomously steered through narrow passages with accuracy and efficiency.

This successful performance was a key highlight of the competition and helped secure the team’s 5th place finish.

February 5th, 2024:

We are currently testing our PID Heading Speed Controller on the RoboBoat side during the first day of the competition. This test focuses on evaluating the performance and responsiveness of the controller as our WAM-V USV navigates through the designated course.

The video showcases the real-time adjustments made by the PID controller to maintain the desired heading and speed, demonstrating its effectiveness in ensuring stable and precise navigation in a competitive environment.

This testing phase is crucial for refining our control algorithms, contributing valuable insights to enhance the overall performance of our autonomous systems as we prepare for the challenges ahead in the RoboBoat competition.

November 10th, 2023:

In this test, we focused on the WAM-V’s autonomous anchor system, designed to deploy the vessel for renewable energy generation. Once anchored, the WAM-V activates a water wheel that harnesses power from the water flow to charge its onboard battery system. This stored energy can then be used to charge a drone, enabling extended autonomous operations.

This system highlights the integration of sustainable energy solutions with autonomous maritime platforms, paving the way for innovative applications in energy harvesting and remote missions.