Optimal Paths for Automated Underwater Vehicles

An MIT team, led by Pierre Lermusiaux, the Doherty Associate Professor in Ocean Utilization, developed a mathematical procedure that can optimize path planning for automated underwater vehicles (AUVs), even in regions with complex shorelines and strong shifting currents. The system can provide paths optimized either for the shortest travel time or for the minimum use of energy, or to maximize the collection of data that is considered most important.Collections of propelled AUVs and gliding AUVs are often used for mapping and oceanographic research, for military reconnaissance and harbor protection, or for deep-sea oil-well maintenance and emergency response.

Transcript

00:00:00 There are a lot of autonomous systems now everywhere in the world and in particular in the ocean we had realized that the number of underwater vehicles is increasing and predicted to increase quite a bit. So, the particular aspects that we are focusing on is two-fold. The first one is trying to estimate the optimal paths in the ocean, in particular focusing on time-optimal, so how to go from point A to point B where you have a lot of underwater vehicles that have all of these paths that they need to plan in an optimal time. Another aspect we are working on is

00:00:34 minimizing energy which becomes equivalent to traveling in optimal time if you travel at a nominal speed for underwater vehicles. One of the amazing things of this project in the sense is really the interdisciplinary character of the work. It involves fluid dynamics and ocean dynamics. It involves advanced numerics and computational schemes. And it involves a deep knowledge of estimation theory and control theory in particular, and also applied mathematics, in the sense of new equations that we have derived. If you think about examples for this new theory and

00:01:12 algorithms that we developed the simplest one is the flow behind and island. So you have an obstacle obviously the underwater vehicles cannot go through the obstacle so they have to avoid it. And they have eddies that form behind the island and so they can utilize these eddies in the most efficient way since the speed of the eddies can be larger at times than that of the vehicle. Similarly if you have underwater vehicles that are released from a point and need to form a formation, lets say of a triangle, and it's behind the exit of a strait, then you have eddies

00:01:47 that form and these vehicles are going to be entrained by the eddies or at times avoid them, or at times go along the flow in order to reach the goal at the end. Another application is if you try to discover ocean processies then specific patterns allow you to extract the dynamics of the flow. Finally you can have regions that are forbidden in the ocean. So in that case it's not really an obstacle, the water goes through the region but you don't want the vehicle to go through that region for many different possible reasons. These could be regions that vehicles would not go either

00:02:26 for a security reason or for pollution reasons etc.. And so therefore we can add those as constraints and those regions can change with time and in this case our algorithm is also capable of dealing with this in an optimal and rigorous, exact way. You could ask the question: Why is this at all important, or; why do we care about traveling in optimal time or minimizing energy; or why do we care to have all of these vehicles underwater? One of the main reasons is that the battery of these underwater vehicles is limited, meaning they cannot travel underwater

00:03:03 forever. There are also questions of bio-falling that can damage the vehicles. So the concept of trying to minimize energy or to travel from one point to another point in optimal time if you travel at nominal speed is very important for many different applications from societal needs to fisheries to protection and security all the way to Naval operation.