Sunday, July 22, 2018
      CMRE Facebook page  CMRE LinkedIn page  CMRE PAO Youtube page
   
Text Size
CMRE banner

Memorandum Reports

Report that covers interim results during the course of a project. Note: Unless linked to the full text, reports are only available to NATO member nations from designated distribution centres.

Documents

Order by : Name | Date | Hits [ Ascendant ]

High-level autonomous asset planner High-level autonomous asset planner

Date added: 10/25/2016
Date modified: 10/25/2016
Filesize: Unknown

High-level autonomous asset planner. Ferri, Gabriele. CMRE-MR-2016-009. October 2016.

In this report we investigate the high level asset management for the autonomous nodes of the CMRE ASW multistatic network. Specifically, our objective is to find a policy to assign the network agents to the tasks which compose an ASW mission. This problem is known as Multi Robot Task Allocation problem in the robotics community. Achieving an effective task allocation is a key driver to fully exploit the benefits of cooperation between the network agents. By "effective" we mean that the allocation scheme, by taking into consideration the evolution of the tactical scene, optimises the team?s objective function by using the agents to the best of their capabilities. A proper task allocation scheme can in fact exploit the synergies that the nodes can offer. For instance, the mobile nodes (robots) can coordinate themselves to patrol a certain area or decide to cooperatively prosecute a track to ease the data fusion process aiding proper target classification. Furthermore the nodes of the network may be heterogeneous, in which case effective task allocation will assign a task to the most suited robot. However, the underwater scenario presents difficult challenges from a task allocation perspective. This is particularly true in littoral waters in which the CMRE network usually operates. In littorals, the typical limitations of underwater communications (low throughput, low range and low reliability) makes the exchange of messages be-tween the robots and between the robots and the C2 centre intermittent and sometimes impossible. A centralised control of the network, which could allow optimal task allocation, is not realisable under these conditions. The network must therefore be capable to allocate the tasks to the different nodes in a distributed way, thereby coping with the severe limitations brought by the underwater acoustic channel. In this report, we propose a market-based approach as a viable solution to task allocation in the ASW scenario. The method works in a completely distributed way and, through periodic auctions, performs the dynamic assignment of robots to tasks during the mission. The algorithm works by using negotiations among neighbouring nodes and requires only local communications. There is no central auctioneer and each robot is able to resolve the current auction and declare the winner by exploiting its local knowledge and the information received by the other nodes. Through periodic auctions, all the robots are sequentially allocated to tasks. In our scenario, we consider continuous tasks (i.e. tasks which do not terminate) consisting of the surveillance of adjacent areas of interest. The proposed utility for the tasks encompasses a measure of the coverage of the area (based on the cumulative probability of detection) within specific frequency constraints. The robots are requested to survey the same area with a certain temporal frequency given the intrinsic dynamic nature of the problem (the target location is dynamically changing). Furthermore, more than one robot can be allocated to the same area introducing cross-schedule dependencies Particular attention has been paid to make the auction scheme robust to intermittent communications. Each robot waits for messages from collaborators for several consecutive communications frames before taking a decision. The periodicity of the auctions makes the scheme robust to the temporary lack of communications. The proposed task allocation policy also allows task reallocations to deal with the evolution of the tactical scene. Results from Matlab simulations are reported and discussed. These results show the robustness of the algorithm to poor communications and how the proposed task al-location policy can improve the network performance by exploiting the cooperation among its nodes.

Optical payloads for unmanned systems Optical payloads for unmanned systems

Date added: 10/25/2016
Date modified: 10/25/2016
Filesize: Unknown

Optical payloads for unmanned systems. Sanjuan Calzado, Violeta. CMRE-MR-2016-008. October 2016.

An optical closure study was developed with a one-of-a-kind optical buoy, designed to acquire simultaneous measurements of the radiant flux field and inherent optical properties. The exercise aims to understand the sources of uncertainty when these measurements are related with the radiative transfer equation. Systematic uncertainties due to sensor inaccuracies were observed together with random uncertainties, due to ambient light variability and different water masses. The results show a departure of up to 70% variability in very clear water. This high variability can potentially have a major impact when optical data is used on cal/val activities and sources of uncertainty should be quantified and presented.

Acoustic sensing of bottom properties using a VLA equipped underwater glider and naturally occurring ambient noise Acoustic sensing of bottom properties using a VLA equipped underwater glider and naturally occurring ambient noise

Date added: 06/22/2016
Date modified: 06/22/2016
Filesize: Unknown

Acoustic sensing of bottom properties using a VLA equipped underwater glider and naturally occurring ambient noise. Nielsen, Peter L. CMRE-MR-2016-005. June 2016.

Seabed properties are critical parameters for sonar performance predictions, and at the same time probably the most difficult information to acquire. Several techniques exist to establish localised geoacoustic properties by probes and advanced inversion schemes, but these methods are often inefficient, require surface vessels and significant human interaction, and are possibly computational intensive. Many of the techniques utilise active sound sources preventing anti-access/denied access operations and limitations in time duration and area coverage. The advancement in development of underwater glider technology combined with hydrophone arrays to measure naturally occurring ambient noise provides a very promising concept to conduct seabed property surveys efficiently. The glider operates almost without human interaction, and the hydrophone array acquires ambient noise to estimate seabed reflectivity and stratification. The ambient noise seabed characterisation has previously been demonstrated successfully using long moored and drifting hydrophone arrays, and the intention is to transition this approach to autonomous vehicles equipped with compact hydrophone arrays. As a part of the development process, CMRE conducted the GLIS-TEN?15 sea trial North of the Elba Island, Mediterranean Sea, where a section of the trial was dedicated to ambient noise measurements. A newly developed prototype survey system composed of a commercial SLOCUM glider equipped with a vertical hydrophone array was deployed to measure ambient noise, while the glider was hovering at depth as part of a series of dive sequences. This procedure samples the seabed properties spatially, and the derived seabed reflection properties and subbottom pro-filing are compared to estimates achieved from two moored reference hydrophone arrays. In general the glider based seabed reflection loss compares favourably to the results from the moored arrays although the loss is slightly lower for the glider data than the moored array data. Only the reflection off the bathymetry was observed from the glider data, while the moored arrays provide detailed seabed layering structure that matches almost perfectly to independent measurements. It is concluded that the glider array data are of poorer quality than the data acquired on the moored arrays, but this does not affect the generality of demonstrated concept.

Underwater acoustic source positioning using a small fleet of underwater gliders equipped with acoustic payload Underwater acoustic source positioning using a small fleet of underwater gliders equipped with acoustic payload

Date added: 06/16/2016
Date modified: 06/16/2016
Filesize: Unknown

Underwater acoustic source positioning using a small fleet of underwater gliders equipped with acoustic payload. Jiang, Yon-Ming, Garau Pujol, Bartolomé. CMRE-MR-2016-003. April 2016.

During the GLISTEN15 (Glider sensors and payloads for tactical characterization of the environment 2015) sea trial, which was conducted North of Elba Italy from 26 August to 09 September 2015, three omni-directional hydrophone-equipped gliders with reactive behaviour capability were deployed as a small glider network for testing its capabilities for maritime acoustic ISR. This report investigates the feasibility of using the fleet of three gliders to localize the signal of interest underwater by means of passive sensing. Two types of source localization algorithms, linearized and nonlinear, were developed and presented in this report. It is identified during the research that the linearized algorithm is suitable for being implemented in gliders for on board processing and for problems in a simple environment, and the nonlinear approach is capable of localizing the signal of interest in all environmental conditions but becomes more computationally demanding. Simulation and preliminary real data analysis showed that it is feasible to localize an acoustic source over 10 km away using a fleet of three gliders. Moreover, simulation results suggested that the algorithms required relative high precision of travel time measurements and also low inconsistency in the data acquisition systems between any two gliders. Hence it is concluded that better synchronization and consistency in data acquisition system of the glider fleet is crucial for applying the algorithms presented in this work to real world data. Further uncertainty analysis should be carried out to understand the quantified requirements, in travel time and glider positions, of the glider system.

Performance of CAS track optimization behavior from LCAS15 Performance of CAS track optimization behavior from LCAS15

Date added: 04/19/2016
Date modified: 04/19/2016
Filesize: Unknown

Performance of CAS track optimization behavior from LCAS15. Ferri, Gabriele. CMRE-MR-2016-004. April 2016.

In this report, a data-driven Mission Management Layer (MML) running on-board AUVs is described. MML manages all the phases of a littoral surveillance mission and exploits the characteristics of Continuous Active Sonar (CAS) signal pro-cessing. The MML selects for further investigation the tracks which are likely related to the presence of a target. In this case, the MML launches a receding horizon, non-myopic control algorithm which controls the AUV?s heading to improve the tracking performance to ease the target classification. The algorithm minimises the expected target position estimation error over a prediction time window by achieving a trade-off amongst different objectives: keeping the target at broadside, reducing the distance to the target, avoiding areas of high reverberation and searching for geometric configurations with low bistatic target localisation error. At-sea experiments performed during the LCAS15 sea trial are reported. They demonstrated, for the first time, that the proposed autonomy architecture can be executed together with real-time Continuous Active Sonar (CAS) processing on-board the AUVs. CAS has recently gained interest for littoral Anti-Submarine Warfare, as it offers the promise of multiple detections per waveform cycle. This can potentially improve the quality/length of tracks, thus increasing the adaptive behaviour?s performance, which, in turn, can increase the detection and tracking capabilities of the processing chain.

User Login