Technology
YOLO-v8 Algorithm
The YOLO Algorithm tells the AutoRescue ship WHERE the person in distress is.
During the training process, the YOLO algorithm was provided with a dataset of over 200 drowning-related images. This algorithm employs a series of neural networks to train a model, which is designed to pinpoint and "box" the location of the person in distress.
During actual rescue operations, real-time images are inputted into the model. The model is anticipated to achieve an accuracy rate of 90.51% in determining the person's location for each image.
Real-Time Movement Algorithm
For all autonomous devices that are required to move, a core movement algorithm that leads the way for the robot is essential. The movement algorithm allows the device to make the best use of the input data collected from the surrounding environment through acceleration sensors and the camera.
It then calculates the corresponding speed for each of the three electric motors so that the vector sum of the forces of the three motors directs the rescue ship to an intended direction.
