1,721,024 research outputs found
Reliability Case Notes No. 9. Autosub Long Range risk assessment report: expert elicitation and risk discussion record
No full textA Risk Assessment Workshop was convened on the 17th November 2014, in Southampton, to quantify the likelihood of loss of Autosub Long Range (ALR) autonomous underwater vehicle in the forthcoming campaign for the UK FASTNEt (Fluxes Across Sloping Topography of the North East Atlantic) science project. The ALR is an autonomous underwater vehicle developed by the National Oceanography Centre, Southampton; its development is funded by the Natural Environment Research Council, UK.This report presents a summary of the discussions, which comprises a review and assessment of potential failure modes and a summary of the risk mitigation actions
Reliability Case Notes No. 10. Board of Inquiry: Circumstances surrounding the stranding of the AutoNaut ‘Gordon’ on the Plymouth coast on 7th November 2014
No full textThe Marine Autonomous Systems in Support of Marine Operations (MASSMO) was a project funded by the Natural Environment Research Council, United Kingdom.Autonaut ‘Gordon’ met all objectives set out in MASSMO mission plan. The vehicle has survived extreme storms in phase 1, whilst gathering valuable science data using state of the art sensors, such as novel passive acoustic devices. In phase 2, the vehicle was successfully deployed to detect tagged fish.In phase 2, Autonaut ‘Gordon’ was deployed on 4th November 2014, at 11.10 am, from Plymouth bay, and transited out 7.2 km East to position 50.337°Latitude, ? 4.13367°Longitude, to start ranging tests and the campaign. The mission was terminated, under NERC’s instructions, on November 6th. During recovery, in the early hours of November 7th, Autonaut ‘Gordon’ was grounded. This resulted in severe damage to the vessel. This Inquiry report looks at the circumstances that have led to the grounding of Autonaut ‘Gordon’ with the aim of diagnosing the most likely root cause for the accident. The grounding of Autonaut ‘Gordon’ was caused by severe weather. However, this report concludes that planning error was the most influential root cause for the accident. The last waypoint, WPT1, was approximately 0.5 miles from Shagstone and lee shore on east side of Harbour entrance. This has put the vehicle in a vulnerable position. The report also concludes that rushed planning was the second most influential root cause for planning error.This report presents analysis of all other potential root causes considered by the Inquiry panel and raises recommendations to mitigate the risk of marine autonomous systems grounding in future deployments
Reliability Case Notes No. 8. Risk and reliability analysis of Autosub 6000 autonomous underwater vehicle
No full textAutosub 6000 is an autonomous underwater vehicle developed and operated by the National Oceanography Centre, Southampton. The vehicle development was funded by the United Kingdom’s Natural Environment Research Council (NERC). The credibility of Autosub 6000 to perform science missions is directly related to its reliability and availability. The aim of this report is to provide reliability and availability analysis of Autosub 6000 based on the vehicle operational history. This report considers the operational history from September 2007 to September 2012, during which seven campaigns were conducted and a total of 68 missions were carried out by Autosub 6000. Eight of the missions ended prematurely.The two most frequently observed failure modes in its fault history were GPS and battery failures. Both failures raised design errors that have since then been addressed.The report concludes that Autosub 6000 failure rate is correlated with two factors: 1. the maximum operating depth and 2. whether or not new technology is being used. Operations at depth greater than 3000 metres are 92% more likely to have an abort than operations at water depth lower than 3000 metres. Adding new technology to the vehicle increases the failure rate by approximately 200%. Of the 110 faults that emerged during the seven campaigns fifty five were mitigated through design corrections, this resulted in a 10% increase in reliability for a 60 km mission, the reliability of which is 0.7.Autosub 6000 software reliability was assessed using two models, the time?related Poisson and the Littlewood. Using both models we can conclude that Autosub 6000 has a reliability of 95% for a 20 hr mission.Lithium?Ion battery reliability was quantified using the FIDES approach. An Autosub 6000 battery pack contains approximately 400 battery cells. The probability of at least one out of four battery packs failing, in 1 year, is 0.20. We concluded that one extra battery pack should be onboard of the vehicle to have at least 96% confidence that the mission will be completed successfully
Analysis of Robustness of Statistical Survival Estimates via Multiple Objective Optimization
No full textReliability Case Notes No. 4: Autosub6000 and Autosub3 actuator potentiometer failure analysis and testing report
Full text linkDuring the James Cook Cruise 27 Autosub6000 aborted mission 12 due to a failure in the position feedback potentiometer of the stern plane actuator. The same actuator is used on Autosub3 which is heading to the Pine Island glacier in Antarctica in January 2009. A similar failure of the Autosub3 actuator while under the ice would result in the loss of the AUV.This report initially describes the investigation into the failure of the feedback potentiometer and shows that the potentiometer’s conductive plastic track became detached from its ceramic substrate and broke up. The report then describes the testing performed to evaluate the reliability of the potentiometer. This involved an accelerated aging test to simulate the worst case conditions seen by the potentiometer in the actuator. This was achieved by oscillating the potentiometer at 4Hz to simulate the actuator movements whilst cycling the pressure of the Morlina 10 oil surrounding the potentiometers.During the testing the 10k? potentiometers used in the actuator were not available, and so 5k? potentiometers from the same range were tested as a substitute. It was assumed that these 5k? potentiometers would produce similar results, however it was found during the testing that the formulation of the 5k? potentiometer track was different from the 10k?; whether this affects the reliability is not known.Due to the large amount of time required to perform each test only 16 5k? potentiometers were tested. Although no failures occurred, the sample was too small to give a high statistical confidence that the potentiometers would survive the cruise. To further reduce the risk four 5k? potentiometers that were to be used on Autosub3 were tested for approximately 72 hours in a ‘burnt in’ process. As an early failure similar to that of Autosub6000 potentiometer would have been detected during this process, the chance of the potentiometers failing was significantly reduced. Thus the burnt in potentiometers were considered acceptable for use on Autosub 3 during the Pine Island campaign
Eliciting expert judgment on the probability of Loss of an AUV operating in four environments
Full text linkAnalysis of causation of loss of communication with marine autonomous systems: a probability tree approach
Full text linkThe last decade has seen the eagerly anticipated introduction of marine autonomous systems as a pragmatic tool for ocean observation. However, outstanding reliability problems means that these vehicles are not yet fulfilling their true potential. Of the classes of problems, loss of communication with a marine autonomous system is both fundamental and difficult to diagnose. In our view, this is due to two reasons: first in many cases users are not technologists and secondly if a vehicle is lost the task of diagnosing the root cause is subject to epistemic uncertainty that users are often reluctant to quantify in a formal manner. As a result users may accept the first hypothesis considered as the main root cause for loss of communication. We show that this approach can result in an increased unreliability of marine autonomous systems through failure to ascertain and then address the true root causes. Consequently, we propose a probability tree approach to help diagnose root cause(s) for loss of communication with a marine autonomous system (MAS). The model was developed based on the results of two detailed investigations and a body of failure data collected from 205 undersea glider operations
On the use of expert judgment elicitation for autonomous underwater vehicle risk prediction and management
No full textAutonomy: Risk Assessment
Full text linkOceanography and ocean observation in general is ever trending toward both automated in situ observation and working in extreme environments. These goals can only be met by de?risking the technology and deployment practices to acceptable levels of risks. A number of industries have standardised risk management processes to support the design and development of their systems. The lack of formal risk assessment of autonomous ocean vehicles has hindered the potential for true autonomy, which is required for exploring unstructured and unexplored environments. When discussing risks different stakeholders may have different consequences foremost in mind. For example the vehicle owner may be interested in risk of loss, whereas the user is interested in risk of vehicle unavailability. Other risks, such as legal risks and risk of collision, affect all stakeholders. This chapter presents a risk management process using several methods tailored to autonomous oceanvehicles in which risk assessment is a key component
Reliability Case Notes No. 7. Risk assessment of the hot-water drilling system for accessing subglacial Lake Ellsworth
No full textThe exploration of Subglacial Lake Ellsworth is a high profile project funded by the Natural Environment Research Council. Clean access to the lake will be provided by a hot-water drilling (HWD) system developed by the British Antarctic Survey. The HWD system is designed to provide a 36 cm diameter borehole through 3.2 km of ice. Drilling to this depth with the HWD system has never been attempted before. This report aims to quantify the risks of the HWD system deployment. A formal assessment of the technical risks is presented. Our analysis was conducted in two parts. First we estimated the probability of failure for all components and processes that take part in each phase of the deployment. In the second part we estimate the availability of the HWD system in light of all potential failure modes. Availability is the probability of the system being available given that it is needed at a given moment.Seventy five potential failure modes have been identified. The assessments for all these failure modes are presented in this report. The probabilities of failure for the top three critical failure modes are:• Boiler failure at 0.048;• Damaged surface hydraulic pipes at 0.0185;• Driller reeler human error at 0.0168.Our availability analysis concluded that, once water circulation has been established, the probability of successfully creating the main hole is 0.83. This will enable the deployment of the probe. Once the main hole has been created, the probability of successfully reaming the main hole, enabling the deployment of the second probe, is 0.89. This is a very high probability of success. These figures take into account the drilling rate and the uncertainty associated with the effectiveness of the drilling process. The current HWD system design has incorporated design changes to avoid previous weaknesses identified during field campaigns. In particular, the current design does not contain hose couplings. This decision has reduced the risk of not achieving the target depth by 15%. <br/
- …
