Mission Brief
During the hackathon our target is to develop a navigation system that:
1. Fuses LiDAR and video to detect and track vessels, buoys, and hazards.
​
2.Uses swarming algorithms for smooth, adaptive waypoint planning.​
​
3. Predicts obstacle motion to adjust course in real time.
​​
4. Ensures reliable navigation in noise, glare, and occlusions
​
​
Background
Autonomous vessels need more than simple waypoint following. DEEPDOCK’s SHARK USV moves safely through busy waters with other ships, buoys, and changing conditions. By combining LiDAR for distance and cameras for object recognition, SHARK clearly understands its surroundings. A nature-inspired control system keeps its motion smooth and safe while avoiding obstacles and staying on course. This reduces risk, prevents collisions, and allows the vessel to operate without constant human control.
Use Case
In coastal waters, SHARK USV autonomously travels to a waypoint while avoiding moored boats, buoys, and a crossing motorboat. Its LiDAR builds a live obstacle map, while the camera feeds an object detection model that labels each object—e.g., “motorboat,” “buoy,” or “sailboat.” The fusion system combines both: LiDAR gives position and distance, vision gives identity.
When a motorboat crosses SHARK’s path, the system predicts its motion and generates a smooth avoidance route that maintains safe distance while staying on course. A nearby buoy is recognized as static, allowing closer navigation. All sensing, classification, prediction, and path planning occur onboard in real time—no operator input required.
registration required !
