Multi-robot Informative Planning for Long-Term Ocean Monitoring

We present an informative path planning method for multiple autonomous underwater vehicles (AUVs) used for long-range and long-term ocean monitoring. We consider the spatio-temporal variations of ocean phenomena, and develop an information-driven approach that computes the most informative observation way-points for reducing the uncertainty for ocean modeling and prediction. The sampling paths of AUVs are then formulated and solved through a matching graph based routing method, which allows the vehicles to transit the obtained informative way-points in an efficient and interference-free way. We provide preliminary simulation results to validate the proposed method.