Autonomous Marine Intelligent Swarming Systems for Interdisciplinary Observing Networks (A-MISSION) - MSEAS Home Page

P.F.J. Lermusiaux, A. Agarwal
P.J. Haley, Jr., T. Sapsis, W.G. Leslie

Massachusetts Institute of Technology
Center for Ocean Engineering
Mechanical Engineering
Cambridge, Massachusetts


Background information
Ongoing MIT research
Presentations
Additional A-MISSION Links

ESSE: (left) Central Forecast; (right) Predicted Optimal Track

Background information

The thrust and scope of our effort is to develop new principled formalisms and methodologies for optimal marine sensing using collaborative swarms of autonomous platforms (AUVs, gliders, ships, moorings and remote sensing platforms) that are smart, i.e. knowledgeable about the predicted environment, acoustic performance and uncertainties, and about the predicted effects of their sensing. Our research focus areas are Autonomous perception and intelligent decision making and Scalable and robust distributed collaboration. Specifically, our work will include research components on: tasking the placement of sensory and computational resources; agile searching; adaptation of algorithms; multi-task learning across multiple sensor types; task allocation, planning and coordination for heterogeneous systems; and structuring autonomy to balance competing tasks. We will also involve the evaluation of uncertain a priori information for decision making; supervisory control of autonomous systems; and methods for acquiring and synthesizing information from multiple sources. Basic automated architectures will also be employed for efficient integration of sensing, planning, and control of autonomous systems.


Ongoing MIT research

Our research will be driven by the following five objectives:

  1. Research autonomous sensing swarms and formations that exploit the multi-scale, multivariate, four-dimensional environmental-acoustic marine dynamics and predictabilities
  2. Utilize swarming schemes based on control theory, dynamical system theory, artificial intelligence and bio-inspired behaviors, and update them so that in the high-level global optimization, data to be collected affect predictions and feedback to the optimal autonomy
  3. Combine the swarming schemes with our adaptive schemes which forecast the impact of future data to define the optimal autonomy
  4. Develop new schemes and compare them step-by-step, in idealized and realistic simulations
  5. Motivate our fundamental research based efficiency and robustness for optimal for naval operations, undersea surveillance, homeland security and coastal protection

Collected research papers are found here.

Return to top of document


Presentations

Return to top of document


Additional A-MISSION Links

Return to top of document


Return to the MSEAS home page