MOOS-DAWG'10

* [[Path:/moos-dawg/material/32-brief-kemna.pdf | Brief given at MOOS-DAWG 2010]]

GLINT09 (29 June to 18 July, 2009) was our first field trial in which adaptive behaviors were tested. A behavior was tested in which an AUV used contact information from on-board real-time processing to adapt its course, keeping a found contact at broad- side to its towed array. For our

GLINT10 field trial (28 July to 16 August, 2010), last year’s work is extended. NURC’s distributed multi-hypothesis tracker is added to the processing chain, enabling the AUV to work on tracks. Furthermore, a behavior was developed that makes the AUV navigate to minimize the localization error of the track, calculated over next possible headings.

%color=#7777BB%[[Talk.32-Kemna|Next-Talk]]%% | \

Within the Cooperative Anti-Submarine Warfare at the NATO Undersea Research Centre, MOOS-IvP is used as the autonomy middleware for our Ocean Explorer au- tonomous underwater vehicles (AUVs). Our end-state demonstration in 2012 will en- compass a multi-static sensor network including cooperative vehicles, distributed in- telligence and distributed processing. This requires our AUVs to be adaptive to their own acquired view of the world, for which IvP behaviors are under development.

This talk discusses the behaviors developed for both field trials, including tests done in simulation and IvP methods used. Furthermore the results of GLINT09, and pre- liminary results of GLINT10, will be presented.

!!!!%color=#449944% '''Stepahnie Kemna, Michael J. Hamilton, David T. Hughes, Robert Been, NATO Undersea Research Centre (NURC), Italy'''

!!!! %color=#7777BB% [[Talk.29-Bays|Prev-Talk]]%%  | \
%color=#7777BB%[[Talk.31-Gagner|Next-Talk]]%% | \

!! Talk-30: ''The Design of a MOOS-IvP Behavior Based on Maximizing the SNR of Autonomous Assets Operating within a Multistatic Sonar System''

!!!!%color=#449944% '''Kevin D. LePage, NATO Underwater Research Center (NURC), Italy'''

The NATO Undersea Research Centre is currently exploring system concepts for autonomous ASW using AUVs. As part of this effort the design of autonomy algorithms (behaviours) which are adaptive on processed sonar signals is being pursued. At the Centre vehicle autonomy is being implemented using MOOS-IvP, an open source software architecture for developing and implementing vehicle autonomy developed at MIT, NUWC, and Oxford University. This is currently used onboard several platforms at NURC, including two Ocean Explorer AUVs. These AUVs each have horizontal line arrays and accompanying signal processing software capable of converting acoustic signals into contacts which can then be acted on by the behaviours implemented within the MOOS-IvP software architecture. In this talk we explore the performance of a behavior which seeks to maximize the future SNR of contacts of interest. This algorithm is based on a model of signal excess and therefore has the ability to seek favorable aspects on the target. Simulations using this algorithm and the resulting emergent vehicle behavior are presented for active ASW scenarios of interest and the performance, advantages, and drawbacks of the approach are discussed.

!!!!%color=#449944% '''Kevin D. LePage, NURC'''

The NATO Undersea Research Centre is currently exploring system concepts for autonomous ASW using AUVs. As part of this effort the design of autonomy algorithms (behaviours) which are adaptive on processed sonar signals is being pursued. At the Centre vehicle autonomy is being implemented using MOOS-IvP, an open source software architecture for developing and implementing vehicle autonomy developed at MIT, NUWC, and Oxford University. This is currently used onboard several platforms at NURC, including two Ocean Explorer AUVs. These AUVs each have horizontal line arrays and accompanying signal processing software capable of converting acoustic signals into contacts which can then be acted on by the behaviours implemented within the MOOS-IvP software architecture. In this talk we explore the performance of a behavior which seeks to maximize the future SNR of contacts of interest. This algorithm is based on a model of signal excess and therefore has the ability to seek favorable aspects on the target. Simulations using this algorithm and the resulting emergent vehicle behavior are presented for active ASW scenarios of interest and the performance, advantages, and drawbacks of the approach are discussed.  Simulation and field test results of the algorithm on a REMUS 100 autonomous underwater vehicle are included.

* Hyrdoid UUVs

%color=#7777BB%[[Talk.Listing|All-Talks]]%% 

!!!!%color=#449944% '''Kevin LePage, NURC'''

!!!! %color=#7777BB% [[Talk.28-Morris|Prev-Talk]]%%  | \
%color=#7777BB%[[Talk.30-Gagner|Next-Talk]]%% | \

!! Talk-29: ''Integration and Testing of a Novel Reacquire/Identify Pattern Generation Algorithm''

!!!!%color=#449944% '''Matthew J. Bays, Jean-François Kamath and Signe A. Redfield, NSWC-PCD'''

We address the integration and field testing of a novel reacquire/identify(RID) pattern generation algorithm.  This algorithm, known as Probabilistic Reacquire/ID Optimal Path Selection (PROPS), is designed to plan a path for a sidescan sonar equipped underwater vehicle in order to produce multiple views of a cluster of discrete targets.  The desired pattern minimizes the total number of turns and time required, while attaining appropriate coverage of the targets. Initial tests of the pattern generation algorithm suggest that it requires between 35% and 95% of the time required by the standard “star” RID pattern.  Following a brief description of the algorithm itself, we present the integration of the algorithm, both as a stand-alone MOOS module and as a library using a standard RID pattern generator created from the MOOS-IvP Helm autonomy toolkit.  Simulation and field test results of the algorithm on a REMUS 100 autonomous underwater vehicle are included.

* MCM

* Navy Labs

!!!! %color=#7777BB% [[Talk.28-Redfield|Prev-Talk]]%%  | \
%color=#7777BB%[[Talk.01-BillinGumstix|Next-Talk]]%% | \

!!!! %color=#7777BB% [[Talk.03-Redfield|Prev-Talk]]%%  | \
%color=#7777BB%[[Talk.05-BillinGumstix|Next-Talk]]%% | \

!!!!%color=#BD614A% '''Categories:''' \

!! Talk-04: ''Integration and Testing of a Novel Reacquire/Identify Pattern Generation Algorithm''

!!!%color=#BD614A% '''Categories:''' \

%color=#BD614A% '''Categories:''' \

%color=#7777BB%[[Talk.05-BillinGumstix|Next-Talk]]%% 

!!!!%color=#449944% '''Matthew J. Bays, Jean- François Kamath and Signe A. Redfield, NSWC-PCD'''

!!!! %color=#7777BB% [[Talk.03-Redfield|Prev-Talk]]%%  |

!!!! %color=#7777BB% [[Talk.03-Redfield|Prev-Talk]]%%  | [[Talk.05-BillinGumstix|Next-Talk]] 

!!!! %color=#444499% [[Talk.03-Redfield|Prev-Talk]]%%  | [[Talk.05-BillinGumstix|Next-Talk]] 

!!!! %color=#449944% [[Talk.03-Redfield|Prev-Talk]]%%  | [[Talk.05-BillinGumstix|Next-Talk]] 

[[Talk.03-Redfield|Prev-Talk]]  | [[Talk.05-BillinGumstix|Next-Talk]] 

[[Talk.04-Redfield|Prev-Talk]]  | [[Talk.04-Redfield|Next-Talk]] 

%color=#449944% '''Matthew J. Bays, Jean- François Kamath and Signe A. Redfield, NSWC-PCD'''

!! ''Integration and Testing of a Novel Reacquire/Identify Pattern Generation Algorithm''

%color=#BD614A% '''Topics:''' \

!! ''MOOS-Enabled Semi-Autonomous Remote USV Operations''

%color=#449944% '''Signe Redfield, NSWC-PCD'''

A multi-vehicle mission involving simultaneous identification (by UUVs) and neutralization (by a USV) of targets is complicated by the need to keep the neutralization efforts distant from the
identification vehicles.  As targets are identified by the UUVs, they are relayed to the USV for imaging (proxy for neutralization).  The USV plans a sequence of neutralization efforts based on desired efficiency (prosecuting targets in close proximity in the same sequence), neutralization capacity (number of targets that can be prosecuted without reloading), the location of the reloading depot, and distance from other vehicles.  We present a solution to this variation of the capacitated vehicle routing problem, implemented on a semi-autonomous USV.  MOOS performed the autonomous portion of the mission running on a remote laptop while a human operator ran a teleoperated underwater vehicle launched and retrieved from the USV as a proxy for the neutralization system as each target was reached. Together the system demonstrated semi-autonomous remote USV operations, with the human operator working smoothly with the autonomous system.

* Neutralization

!! ''Autonomous Adaptive Environmental Feature Tracking on Board AUVs: Tracking the Thermocline''

%color=#449944% '''Stephanie Petillo, MIT (LAMSS)'''

This talk addresses the challenge of autonomously and adaptively tracking features of the underwater environment using AUVs running the MOOS-IvP autonomy software.  This problem is addressed from concept to implementation in the field on various AUV platforms, developing specifically the example of thermocline tracking.  Some recent research involving methods for feature tracking on board multiple AUVs operating simultaneously and collaboratively to detect an underwater feature will also be discussed briefly.

One of the greatest challenges of working in the underwater regime is the severe limitations of acoustic communications. This problem becomes even more evident in multi-vehicle autonomy, when vehicles must continually update each other with their state and intentions to achieve cooperative goals. In order to support tests of a multi-vehicle arbiter framework, an optimization scheme was created and implemented as a MOOS module to enable sufficient message passing between vehicles. Using this tool, vehicle state and destination, shared map updates, updated algorithm parameters, target information, and decision reconciliation can be effectively shared between vehicles using the published Compact Control Language (CCL) standard for acoustic messages.

* Acoustic Communications,

* Autonomy%%

%color=#449944% '''Stephanie Petillo, MIT (LAMSS)''

!! ''Unmanned Robot Message Optimization Method (URMOM)''

%color=#449944% '''Andrew Bouchard, NSWC-PCD'''

%color=#BD614A% '''Topics:''' Acoustic Communications, Multi-Vehicle Autonomy, Autonomy%%

''Topics:'' Acoustic Communications, Multi-Vehicle Autonomy, Autonomy

One of the greatest challenges of working in the underwater regime is the severe limitations of acoustic communications. This problem becomes even more evident in multi-vehicle autonomy, when vehicles must continually update each other with their state and intentions to achieve cooperative goals. In order to support tests of a multi-vehicle arbiter framework, an optimization scheme was created and implemented as a MOOS module to enable sufficient message passing between vehicles. Using this tool, vehicle state and destination, shared map updates, updated algorithm parameters, target information, and decision reconciliation can be effectively shared between vehicles using the published Compact Control Language (CCL) standard for acoustic messages.

%color=#449944% '''Andrew Bouchard, NSWC PCD'''

%color=#449944% '''Andrew Bouchard, NSWC PCD'''%

%color=#449944% !!! '''Andrew Bouchard, NSWC PCD'''%

!!! '''Andrew Bouchard, NSWC PCD'''

'''Andrew Bouchard, NSWC PCD'''

!! Title: ''Unmanned Robot Message Optimization Method (URMOM)''

Title: ''Unmanned Robot Message Optimization Method (URMOM)''

Title: Unmanned Robot Message Optimization Method (URMOM)

Andrew Bouchard, NSWC PCD

     One of the greatest challenges of working in the underwater regime is the
severe limitations of acoustic communications. This problem becomes even more evident in multi-vehicle autonomy, when vehicles must continually update each other with their state and intentions to achieve cooperative goals. In order to support tests of a multi-vehicle arbiter framework, an optimization scheme was created and implemented as a MOOS module to enable sufficient message passing between vehicles. Using this tool, vehicle state and destination, shared map updates, updated algorithm parameters, target information, and decision reconciliation can be effectively shared between vehicles using the published Compact Control Language (CCL) standard for acoustic messages.

     One of the greatest challenges of working in the underwater regime is the \
severe limitations of acoustic communications. This problem becomes even more e\
vident in multi-vehicle autonomy, when vehicles must continually update each ot\
her with their state and intentions to achieve cooperative goals. In order to s\
upport tests of a multi-vehicle arbiter framework, an optimization scheme was c\
reated and implemented as a MOOS module to enable sufficient message passing be\
tween vehicles. Using this tool, vehicle state and destination, shared map upda\
tes, updated algorithm parameters, target information, and decision reconciliat\
ion can be effectively shared between vehicles using the published Compact Cont\
rol Language (CCL) standard for acoustic messages.