Fig 5 - uploaded by Enrico Anderlini
Content may be subject to copyright.
Free-body diagram showing the forces acting on the UG in the vertical plane in steady-state conditions.
Source publication
Marine growth has been observed to cause a drop in the horizontal and vertical velocities of underwater gliders, thus making them unresponsive and needing immediate recovery. Currently, no strategies exist to correctly identify the onset of marine growth for gliders and only limited datasets of biofouled hulls exist. Here, a field test has been run...
Contexts in source publication
Context 1
... free-body diagram of the equilibrium condition for the steady-state flight is shown in Fig. 5a and Fig. 5b for descents and ascents, respectively. B indicates the net buoyancy, L the lift and D the drag force. U is the surge velocity component in the body-fixed frame, θ the pitch, α the attack and β the glide-path angles. The glide-path angle indicates the angle of the flight path in the inertial reference system and is obtained ...
Context 2
... free-body diagram of the equilibrium condition for the steady-state flight is shown in Fig. 5a and Fig. 5b for descents and ascents, respectively. B indicates the net buoyancy, L the lift and D the drag force. U is the surge velocity component in the body-fixed frame, θ the pitch, α the attack and β the glide-path angles. The glide-path angle indicates the angle of the flight path in the inertial reference system and is obtained from the ...
Similar publications
![Figure 1.1: A HDPE gravity cage [4] (a), an interconnected hinged steel...](profile/Farshad-Nobakht-Kolur-2/publication/349724123/figure/fig1/AS:996987755319298@1614711812805/A-HDPE-gravity-cage-4-a-an-interconnected-hinged-steel-fish-farm-5-b-and-a_Q320.jpg)
Ever growing global demand for aquatic resources due to population growth, limitation in freshwater resources and uncertainty and restrictions associated with wild fish stocks and other aquatic resources have pushed the aquaculture industry offshore into deeper waters. Concurrent advances in the offshore technology provided a ground for the develop...
Citations
... Apart from the three areas mentioned in this paper, there are several other areas in the marine field that have been facilitated by ML such as underwater vehicles (Anderlini et al., 2020b, wave energy converters (Anderlini et al., 2018b;Benites-Munoz et al., 2020), condition monitoring and maintenance (Adeli and Kim, 2004;Kim and Adeli, 2005;Amezquita-Sanchez and Adeli, 2015;Amezquita-Sanchez et al., 2017;Perez-Ramirez et al., 2016;Li et al., 2020b). ...
The shipping industry faces a large challenge as it needs to significantly lower the amounts of Green House Gas emissions. Traditionally, reducing the fuel consumption for ships has been achieved during the design stage and, after building a ship, through optimisation of ship operations. In recent years, ship efficiency improvements using Machine Learning (ML) methods are quickly progressing, facilitated by available data from remote sensing, experiments and high-fidelity simulations. The data have been successfully applied to extract intricate empirical rules that can reduce emissions thereby helping achieve green shipping. This article presents an overview of applying ML techniques to enhance ships' sustainability. The work covers the ML fundamentals and applications in relevant areas: ship design, operational performance, and voyage planning. Suitable ML approaches are analysed and compared on a scenario basis, with their space for improvements also discussed. Meanwhile, a reminder is given that ML has many inherent uncertainties and hence should be used with caution.
... The axisymmetric low drag shape of Seaglider can maintain laminar flow over more than 80% of its surface even at high speeds that can reach 7m/sec [8]. Gliders, never operate at such high speeds since in general, they are slow in their nature compared to AUVs, which makes them even more susceptible to biofouling [10]. It is worth mentioning that Seaglider had a conductivity sensor on its nose that had only 2% of the cross-sectional area of the glider's nose. ...
p> This paper outlines the design and testing process of the hull of a deep small Autonomous Underwater Vehicle (AUV), rated at 2000m depth. Many existing AUV pressure housings use aluminum or other isotropic traditional metals, instead of composites due to the complexities of the design of composites at such big load. The research at hand explains the process of design starting from setting the geometrical constraints for the design to mass production. To the best of the authors’ knowledge, none of the previous studies has presented such detailed description of the work. Carbon fiber reinforced epoxy material was chosen thanks to its high strength-to-weight ratio and similarity of its compressibility to sea water. Material characterization was performed to obtain the material properties under loading conditions using a modified method of the Combined Loading Compression testing technique. A specific fixture was designed and manufactured to test filament-wound tubes. An analytical model was developed using MATLAB, a finite element model was created using ABAQUS, and the results of the two models were compared. A set of recommendations was introduced for the stacking sequence to provide the lowest possible stresses, regardless on the diving depth of the vehicle. Afterwards, a quality control set of tests was conducted, including seawater absorption under high pressure and void analysis using destructive and non-destructive tests. Pilot samples were manufactured and tested in a pressure vessel, where it was cycle-tested and inspected using visual and ultrasonic testing. Other samples were fail-tested and showed a failure at ∼93% of the expected failure load. Such range can be considered good to provide safe operation for the vehicle at the designated depth, given that the factor of safety included covers more than 7% of the failure load. The proposed design methodology has shown that CFRE can be safely used even at such high depths. </p
... The axisymmetric low drag shape of Seaglider can maintain laminar flow over more than 80% of its surface even at high speeds that can reach 7m/sec [8]. Gliders, never operate at such high speeds since in general, they are slow in their nature compared to AUVs, which makes them even more susceptible to biofouling [10]. It is worth mentioning that Seaglider had a conductivity sensor on its nose that had only 2% of the cross-sectional area of the glider's nose. ...
p>This paper presents the design and testing process of the hull of a deep small autonomous underwater vehicle, rated at 2000m depth. Starting from setting the geometrical constraints for the design to mass production. None of the previous literature, to the authors’ knowledge has presented similar work. Carbon Fiber Reinforced Epoxy material was chosen given their high strength to weight ratio and similarity of its compressibility to sea water. Material characterization was performed to obtain the material properties under loading conditions using a modified method of the Combined Loading Compression testing technique. A specific fixture was designed and manufactured to test filament-wound tubes. An analytical model was developed on MATLAB, a finite element model was created on ABAQUS, and the results of both models were compared. Then a quality control set of tests was performed including seawater absorption under high pressure and void analysis using destructive and non-destructive tests. Pilot samples were manufactured and tested in a pressure vessel, where one was cycle tested and inspected using visual and ultrasonic testing, and the other one was fail tested where it failed at ~93% of the expected load.</p
... Frajka-Williams et al. (2011) discovered failures in a pitch tilt sensor. Furthermore, biofouling was studied in Haldeman et al. (2016), Eichhorn et al. (2020), Anderlini et al. (2020b) and wing loss in Anderlini et al. (2020a). Marine growth on a gliders hull can greatly increase its weight and significantly reduce its speed, presenting a large operational issue that can require premature retrieval at sea. ...
... Marine growth on a gliders hull can greatly increase its weight and significantly reduce its speed, presenting a large operational issue that can require premature retrieval at sea. Severe biofouling has been seen to increase the cost of transport by 63.3% and lower horizontal speed by 46.3% (Anderlini et al., 2020b). As opposed to marine growth, which typically spans many weeks, wing loss can occur suddenly and may also go unnoticed by the pilot until the glider is retrieved (Anderlini et al., 2020a). ...
... A steady-state dynamic model is adopted by Eichhorn et al. (2020) to investigated the change in dynamic parameters over the duration of a dive. As detailed in Anderlini et al. (2020a), this solution shows distinct changes in the model parameters for wing loss, or in case study simulating severe biofouling conditions (Anderlini et al., 2020b). Again, the system was effective, however, there is a great upfront computational cost to the global optimisation of dynamic constants. ...
Autonomous underwater vehicles (AUVs) are used extensively for monitoring the world's oceans, taking measurements of oceanographic characteristics along the water column. Presently, there is no holistic anomaly detection system in operation and AUVs require experienced pilots to monitor the progress of missions. This results in a large operational overhead and reduces the number of AUVs that can be deployed simultaneously. This article proposes an online anomaly detection system for underwater gliders based on a data-driven approach. A novel Long Short-Term Memory (LSTM) Variational Autoencoder (VAE) has been developed and trained using field data from four deployments with healthy glider behaviour and then tested against four deployments where faults are present. The system is able to detect wing loss with a high degree of accuracy on gliders unseen during the models training, highlighting the generality of the model to different platforms. Additionally, the VAE method outperforms model-based solution for the detection of biofouling, proving its generality to different types of anomalies. The proposed smart anomaly detection will contribute to increasing the capacity of AUVs and reducing the dependence on support vessels and experienced pilots.
... The axisymmetric low drag shape of Seaglider can maintain laminar flow over more than 80% of its surface even at high speeds that can reach 7m/sec [8]. Gliders, never operate at such high speeds since in general, they are slow in their nature compared to AUVs, which makes them even more susceptible to biofouling [10]. It is worth mentioning that Seaglider had a conductivity sensor on its nose that had only 2% of the cross-sectional area of the glider's nose. ...
This paper outlines the design and testing process of the hull of a deep small Autonomous Underwater Vehicle (AUV), rated at 2000m depth. Many existing AUV pressure housings use aluminum or other isotropic traditional metals, instead of composites due to the complexities of the design of composites at such big load. The research at hand explains the process of design starting from setting the geometrical constraints for the design to mass production. To the best of the authors’ knowledge, none of the previous studies has presented such detailed description of the work. Carbon fiber reinforced epoxy material was chosen thanks to its high strength-to-weight ratio and similarity of its compressibility to sea water. Material characterization was performed to obtain the material properties under loading conditions using a modified method of the Combined Loading Compression testing technique. A specific fixture was designed and manufactured to test filament-wound tubes. An analytical model was developed using MATLAB, a finite element model was created using ABAQUS, and the results of the two models were compared. A set of recommendations was introduced for the stacking sequence to provide the lowest possible stresses, regardless on the diving depth of the vehicle. Afterwards, a quality control set of tests was conducted, including seawater absorption under high pressure and void analysis using destructive and non-destructive tests. Pilot samples were manufactured and tested in a pressure vessel, where it was cycle-tested and inspected using visual and ultrasonic testing. Other samples were fail-tested and showed a failure at ~93% of the expected failure load. Such range can be considered good to provide safe operation for the vehicle at the designated depth, given that the factor of safety included covers more than 7% of the failure load. The proposed design methodology has shown that CFRE can be safely used even at such high depths.
... In [3], a simple but effective system was developed to detect the wing loss using the roll angle. In [4], system identification techniques were employed to detect changes in model parameters which further successfully deduced simulated and natural marine growth. Anderlini, et al. [5] further conducted a field test to validate a marine growth detection system for UGs using ensembles of regression trees. ...
... Hamilton et al. (2007) propose an integrated fault detection and diagnosis architecture for AUVs, although the focus is on on-board systems. Anderlini et al. (2020a) have designed rule-and model-based methods for the detection of the loss of wings on UGs and the onset of high levels of marine growth on UGs (Anderlini et al., 2020b). Further work by Anderlini et al. (2021a) has developed and tested an anomaly detection system that blends model-and data-based solutions to detect both simulated and naturally accumulated biofouling. ...
... due to ocean currents, and bulky sensory packs, e.g. the turbulence probes for the Ocean Microstructure Gliders (OMG). Biofouling caused by marine growth in shallow, warm and tropical waters can lead to an increase of a UG's weight, a significant drop in speed, even possible premature retrieval at sea (Anderlini et al., 2020b). OMG would lead to a higher drag coefficient and a higher negative buoyancy offset. ...
... The anomaly score increases gradually from dive cycle 200, which is likely associated with marine growth (Haldeman et al., 2016). The growing anomaly score is in line with the increasing drag coefficient deduced through model-based and other data-driven approaches in Anderlini et al. (2020b). The similar growth of the anomaly score and the drag coefficient suggest that the proposed BiGAN-based anomaly detection system can capture slowly growing anomalies, even though it is trained with deployment datasets collected by other gliders in different missions. ...
An effective anomaly detection system is critical for marine autonomous systems operating in complex and dynamic marine environments to reduce operational costs and achieve concurrent large-scale fleet deployments. However, developing an automated fault detection system remains challenging for several reasons including limited data transmission via satellite services. Currently, most anomaly detection for marine autonomous systems, such as underwater gliders, rely on intensive analysis by pilots. This study proposes an unsupervised anomaly detection system using bidirectional generative adversarial networks guided by assistive hints for marine autonomous systems with time series data collected by multiple sensors. In this study, the anomaly detection system for a fleet of underwater gliders is trained on two healthy deployment datasets and tested on other nine deployment datasets collected by a selection of vehicles operating in a range of locations and environmental conditions. The system is successfully applied to detect anomalies in the nine test deployments, which include several different types of anomalies as well as healthy behaviour. Also, a sensitivity study of the data decimation settings suggests the proposed system is robust for Near Real-Time anomaly detection for underwater gliders.
... Leaks, motor or pump malfunctions, low-battery voltage and sensor dropouts or faults are also identified as problems in Schofield et al. (2007) Additionally, Frajka-Williams et al. (2011) have observed the failure of the pitch tilt sensor. Furthermore, the authors have investigated the loss of wing for UGs in Anderlini et al. (2020a) and the impact of biofouling in Anderlini et al. (2020b). ...
... Additionally, for units 304 and 436 the rule-and model-based measures to detect wing loss from Anderlini et al. (2020a) are included, namely the difference in the mean roll angle in ascents and descents and the buoyancy offset, respectively. For unit 399, the drag coefficient is the model-based metric to detect marine growth as in Anderlini et al. (2020b). ...
Marine Autonomous Systems (MAS) operating at sea beyond visual line of sight need to be self-reliant, as any malfunction could lead to loss or pose a risk to other sea users. In the absence of fully automated on-board control and fault detection tools, MAS are piloted and monitored by experts, resulting in high operational costs and limiting the scale of observational fleets that can be deployed simultaneously. Hence, an effective anomaly detection system is fundamental to increase fleet capacity and reliability. In this study, an on-line, remote fault detection system is developed for underwater gliders. Two alternative methods are analysed using time series data: feedforward deep neural networks estimating the glider's vertical velocity and an autoencoder. The systems are trained using field data from four baseline deployments of Slocum gliders and tested on six deployments of vehicles suffering from adverse behaviour. The methods are able to successfully detect a range of anomalies in the near real time data streams, whilst being able to generalise to different glider configurations. The autoencoder's error in reconstructing the original signals is the clearest indicator of anomalies. Thus, the autoencoder is a prime candidate to be included into an all-encompassing condition monitoring system for MAS.
... The nearest neighbor classifier was found to be especially highly accurate over two different test sets. In [18] and [19], model-based methods were used to detect the loss of a wing and high levels of marine growth specifically for UGs. Here, the second study is extended with the design of data-driven solutions for the detection and isolation of marine growth on UGs. ...
Marine growth has been observed to cause a drop in the horizontal and vertical velocities of underwater gliders, thus making them unresponsive and needing immediate recovery. Currently, no strategies exist to correctly identify the onset of marine growth for gliders and only limited data sets of biofouled hulls exist. Here, a field test has been conducted to first investigate the impact of marine growth on the dynamics and power consumption of underwater gliders and then design an anomaly detection system for high levels of biofouling. A Slocum glider was deployed first for eight days with drag stimulators to imitate severe biofouling; then, the vehicle was redeployed with no additions to the hull for further 20 days. The mimicked biofouling caused a speed reduction due to a significant increase in drag. Additionally, the lower speed causes the steady-state flight stage to last longer and the rudder to become less responsive; hence, marine growth results in a shortening of deployment duration through an increase in power consumption. As actual biofouling due to
p. pollicipes
occurred during the second deployment, it is possible to develop and test a system that successfully detects and identifies high levels of marine growth on the glider, blending model- and data-based solutions using steady-state flight data. The system will greatly help pilots replan missions to safely recover the vehicle if significant biofouling is detected.
... aircraft [13]. The authors successfully introduced an effective model-based system for the detection and isolation of the loss of one wing [12] and high levels of biofouling [14] and on underwater gliders. ...
ALADDIN Extended Abstract The Assuring Autonomy International Programme (AAIP) has the ambitious goal of delivering a Body of Knowledge (BoK) for the assurance of Robotics and Autonomous Systems (RAS). The demonstrator project "Assuring Long-term Autonomy through Detection and Diagnosis of Irregularities in Normal operation (ALADDIN)" (including project partners University College London (UCL) and National Oceanography Centre (NOC) and project stakeholders Lloyd's Register (LR), Maritime and Coastguard Agency (MCA) and Blue Ocean Monitoring (BOM)) will contribute to the drafting of the sections on the definition and verification of sensing and understanding requirements for RAS and the identification of sensing (and understanding) deviations. To achieve this, new methods for the detection and identification of adverse behaviour for Marine Autonomous Systems (MAS) will be introduced. Then, they will be implemented on the command-control infrastructure for over-the-horizon operations of MAS (C2)-itself a RAS-developed by the NOC for verification. This technical report represents the first deliverable of the project and concludes Work Package (WP) 1. The aim of WP1 was to define healthy and anomalous behaviour of RAS with appropriate vocabulary, identify data availability and sources for RAS deployments, with a focus on MAS, and outline the project steering direction for the following WPs. In the absence of accepted international terminology for MAS at present, the vocabulary for condition monitoring of MAS used in project ALADDIN has been introduced. The terminology is designed to be general and transferable to any RAS technology, although failure modes are specific to each design. The selected procedure for the robust labelling of RAS timeseries data consists in first defining baseline standard RAS behaviour, then tagging anomalous behaviour in a single class and finally characterising the anomalous behaviour according to the specific fault. If unsupervised learning methods are used to automatically label the data, a final layer is included to enable human feedback and correcting erroneous tagging. Metadata based on the timestamp is exploited to help the identification of unhealthy RAS conditions on the levels of a whole RAS deployment, segments of the deployment and individual operational cycles. Furthermore, specific datasets of deployments of a range of Autonomous Underwater Vehicle (AUV) types have been selected to be used in the remainder of the project, enabling the design of solutions that are transferable to different technologies. The datasets will be complemented by remote-sensing (satellite data) to capture the environmental disturbances during the deployments. After identifying an important research gap, project ALADDIN will focus on multi-platform sensing and deep learning for fleet management fault diagnostics to maximise its contribution to advancing the state of the art of intelligent maintenance systems. At present, a generative-adversarial-network-based method is being introduced that will be used for the definition of sensing requirements and deviations of RAS in WP2. The method will be improved with a hybrid model-based and data-driven architecture in WP3 to define understanding requirements and deviations. Additionally, code is being implemented on C2 for the verification of the methods in WP4, with some field tests of AUVs being planned.
