//==================================================================== module = lib_henrik_anneal type = Library author = Henrik Schmidt contact = henrik@mit.edu org = MIT thumb = A library with annealer and thermal front sim, for 2.680 Front Estimation lab depends = borndate = 120301 doc_url = http://oceanai.mit.edu/ivpman/labs/front_estimate license = GPL //group = MOOS-IvP, Libraries, 2.680, Simulation group = 2.680 distro = moos-ivp.org synopsis = The front estimation lab utilizes a simulated moving thermal front. The simulator is implemented in this library to support both the actual simulator MOOS app (uFldCTDSensor) and the MOOS app generating information for rendering during simulation (uFrontGridRender). This library also provides a strawman MOOS app for estimating the thermal front parameters using a simulated annealer. The annealer is contained in this library to support both the strawman app (pFrontEstimate), and any student written app written during the lab. //==================================================================== module = lib_ufld_hazards type = Library author = Michael Benjamin contact = mikerb@mit.edu, issues@mit.edu org = MIT thumb = A library containing key data structures for the 2.680 Hazard Search Lab depends = borndate = 120112 doc_url = http://oceanai.mit.edu/ivpman/apps/uFldHazardSensor license = GPL //group = MOOS-IvP, Libraries, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The lib_ufld_hazards library contains key classes used by the handful of applications used in the Hazard Search lab of MIT 2.680. It holds the class definition of a Hazard object as well as methods for serialization and de-serialization. //==================================================================== module = gen_hazards type = Command line utility author = Michael Benjamin contact = mikerb@mit.edu org = MIT thumb = A utility for generating a hazard file depends = lib_mbutil, lib_ufld_hazards, geometry borndate = 120127 doc_url = Pending license = GPL group = 2.680 distro = moos-ivp.org synopsis = gen_hazards is a utility for generating a hazard file, i.e., a simulated threat laydown. Objects will be generated within a given convex polygon. Generated objects are either hazards or benign objects. Benign objects may have a resemblance factor associated with them. This represents how similar a benign object may appear to a hazardous object. Objects may also have an optimal sensing aspect specified. This represents the best approach angle for a vehicle to optimize its sensor performance. If the optimal aspect is specified, a sensitivity factor is also given, ssigma. //==================================================================== module = pFrontEstimate type = MOOS App author = Henrik Schmidt contact = henrik@mit.edu org = MIT thumb = Evualation of a simulated thermal front depends = lib_mbutil, lib_henrik_anneal, lib_MOOS borndate = 120315 doc_url = Pending license = GPL //group = MOOS-IvP, Simulation, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The pFrontEstimate app is a baseline implementation of an app to be written as part of the 2.680 Front Estimation lab. Those doing this lab will typically start with this app and improve it. //==================================================================== module = pFrontGridRender type = MOOS App author = Michael Benjamin, Henrik Schmidt contact = mikerb@mit.edu org = MIT thumb = Renders a simulated moving thermal front depends = lib_mbutil, lib_geometry, lib_henrik_anneal, lib_MOOS borndate = 130413 doc_url = Pending license = GPL //group = MOOS-IvP, Simulation, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The pFrontGridRender is typically used in the Front Estimation lab of MIT 2.680, to render the thermal front currently being simulated. This app typically runs on the shoreside computer during the in-water lab. The simulation of the front itself is typically done in uSimCTDSensor, which posts the present temperature values associated with a x-y coordinates. pFrontGridRender maintains a grid structure, and posts viewable objects to be consumed by pMarineViewer for rendering. //==================================================================== module = pGradeFrontEstimate type = MOOS App author = Henrik Schmidt, Michael Benjamin contact = mikerb@mit.edu org = MIT thumb = Run-time evaluation of an estimate of simulated thermal front depends = lib_mbutil, lib_geometry, lib_henrik_anneal, lib_MOOS borndate = 120301 doc_url = Pending license = GPL //group = MOOS-IvP, Simulation, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The pGradeFrontEstimate application is part of 2.680 Front Estimation lab. It typically runs on the shoreside computer and receives incoming estimates of the wave front, sent by vehicle exploring the operation area. It has access to the ground truth, actaual wave front parameters, used by uSimCTDSensor. These parameters come via the UCTD_TRUE_PARAMETERS published by uSimCTDSensor. //==================================================================== module = uFldCTDSensor type = MOOS App author = Henrik Schmidt, Michael Benjamin contact = mikerb@mit.edu org = MIT thumb = Simulates a CTD Sensor and simulated thermal front depends = lib_mbutil, lib_apputil, lib_geometry, lib_henrik_anneal, lib_MOOS borndate = 120128 doc_url = Pending license = GPL //group = MOOS-IvP, Simulation, uField ToolBox, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The uFldCTDSensor was originally written to the support the MIT 2.680 Front Estimation lab. It is typically run in a shoreside community. Simulates a CTD sensor on a vehicle. Currently only provides temperature. Configured for a specific temepraturte field, currently a propagating ocean front. //==================================================================== module = uFldHazardMetric type = MOOS App author = Michael Benjamin contact = mikerb@mit.edu org = MIT thumb = Receives and grades effectiveness of hazard field reports depends = lib_mbutil, lib_apputil, lib_geometry, lib_ufld_hazards, lib_MOOS borndate = 120312 doc_url = http://oceanai.mit.edu/ivpman/apps/uFldHazardMetric license = GPL //group = MOOS-IvP, Simulation, uField ToolBox, 2.680 group = 2.680 distro = moos-ivp.org synopsis = A utility for quickly evaluating a hazardset report; a list of declared hazards and their locations. Evaluating a hazardset report against ground truth and a reward structure is fairly straight-forward, but tedious. This tool performs this operation automatically, and as a MOOS process with the result posted both to the MOOSDB and viewable in the appcast output of uFldHazardMetric. Operation is comprised of a few simple parts: (1) Import a ground-truth hazard field: A ground truth hazard field is a text file listing the location of hazards and hazard-like objects, and their locations. This file also typically includes a search region, a convex polygon containing all listed objects. A uFldHazardMetric configuration parameter names the file. (2) Import a reward structure: A reward structure, consisting of penalties for missed hazards and false alarms, is imported as a set of uFldHazardMetric configuration parameter. (3) Evaluate a hazardset report: A hazardset report is received by MOOS mail and evaluated item by item against the ground truth and reward structure. The results are then posted and rendered. This step is repeated for each received report. //==================================================================== module = uFldHazardMgr type = MOOS App author = Michael Benjamin contact = mikerb@mit.edu org = MIT thumb = Strawman MOOS app for managing hazard sensor information depends = lib_mbutil, lib_apputil, lib_geometry, lib_ufld_hazards, lib_MOOS borndate = 121026 doc_url = http://oceanai.mit.edu/ivpman/apps/uFldHazardMgr license = GPL //group = MOOS-IvP, Simulation, uField ToolBox, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The uFldHazardMgr is a strawman MOOS app for managing hazard sensor information and generation of a hazard report over the course of an autonomous search mission. In the 2.680 Hazard Search lab, this module is regarded as a baseline solution to be replaced by a more sophisticated solution by those doing the lab. //==================================================================== module = uFldHazardSensor type = MOOS App author = Michael Benjamin contact = mikerb@mit.edu org = MIT thumb = Simultated (e.g. side scan) sensor for detecting underwater objects depends = lib_mbutil, lib_apputil, lib_geometry, lib_ufld_hazards, lib_MOOS borndate = 120128 doc_url = http://oceanai.mit.edu/ivpman/apps/uFldHazardSensor license = GPL //group = MOOS-IvP, Simulation, uField ToolBox, 2.680 group = 2.680 distro = moos-ivp.org synopsis = The uFldHazardSensor MOOS app simulates a sensor for detecting underwater objects. It is typically run in a shoreside community. Configured with a set objects (i.e., "threat laydown" if the objects of interest are mines) with a given x,y location and classification (hazard or benign). The sensor simulator receives a series of requests from a remote vehicle. When sensor determines that an object is within the sensor field of a requesting vehicle, it may or may not return a sensor detection report for the object, and perhaps also a proper classification. The odds of receiving a detection and proper classification depend on the sensor configuration and the user's preference for P_D/P_FA on the prevailing ROC curve.