This paper addresses the problem of horizon detection, a fundamental process in numerous object detection algorithms, in a maritime environment. The maritime environment is characterized by the absence of fixed features, the presence of numerous linear features in dynamically changing objects and background and constantly varying illumination, rendering the typically simple problem of detecting the horizon a challenging one. We present a novel method called multi-scale consistence of weighted edge Radon transform, abbreviated as MuSCoWERT. It detects the long linear features consistent over multiple scales using multi-scale median filtering of the image followed by Radon transform on a weighted edge map and computing the histogram of the detected linear features. We show that MuSCoWERT has excellent performance, better than seven other contemporary methods, for 84 challenging maritime videos, containing over 33,000 frames, and captured using visible range and near-infrared range sensors mounted onboard, onshore, or on floating buoys. It has a median error of about 2 pixels (less than 0.2%) from the center of the actual horizon and a median angular error of less than 0.4 deg. We are also sharing a new challenging horizon detection dataset of 65 videos of visible, infrared cameras for onshore and onboard ship camera placement.
The Automatic Identification System (AIS) tracks vessel movement by means of electronic exchange of navigation data between vessels, with onboard transceiver, terrestrial and/or satellite base stations. The gathered data contains a wealth of information useful for maritime safety, security and efficiency. This paper surveys AIS data sources and relevant aspects of navigation in which such data is or could be exploited for safety of seafaring, namely traffic anomaly detection, route estimation, collision prediction and path planning.
The role of marine engineers is to maintain the operational state of all systems in the engine room such as to diagnose and rectify problems arising, and to understand what maintenance will be required to keep the vessel appropriately operational and safe. Through training and experience, the engineering crew can read and interpret engine room instrumentation, and employ their intuitive feel for normal operation in-situ. In order to examine the possibilities of future remote and autonomous uncrewed vessels, we developed a virtual reality simulated engine room based on a real vessel. Even though the end product is effectively a virtual simulation, the original audio from the engine room was recorded and used, thus providing a more accurate and immersive experience to the users. This paper examines the use and application of a remote server to feed audio and other data in the simulated virtual engine and to create hypothetical failures scenarios for testing and training. Experience engineers carried out different failure scenarios as they usually do onboard and provided valuable feedback. User testing suggests that upcoming paradigms of the Internet of Audio Things can become a vital element in future operations of Maritime Autonomous Surface Ships.
Safety at sea is the protection from harm to people, property and the environment. Safety assurance in the case of autonomous sea going vessels is nontrivial due to the pace of change in enabling technologies and their disruptive impact. Historically accidents and incidents at sea have often been attributed to human error but the safety implications of a machine rather than a human making decisions whether fully or in part, is yet to be understood. Although the development of regulation of autonomy at sea is in its early stages, there is much activity to address safety of autonomy in maritime and elsewhere, along with a wealth of established safety practice from before its advent with good read across. Integral to the development of a safety case is the assessment and mitigation of risk i.e. the combination of frequency and severity of the consequence. When assessed quantitatively, frequency historically has been calculated using failure rates of components whose failure leads to the materialisation of the risk. These failure rates historically were derived from mechanical component failures, which were relatively easy to determine. Recently, the advent of software-controlled systems has introduced difficulties in defining failure rates; an issue which is exacerbated greatly in the context of autonomy, and within complex autonomous systems it is not possible to perform quantitative risk assessments using failure rates. This gives rise to a need to use different techniques to assess risk in the development of a maritime autonomy related safety case. Two recent and significant developments are the EU H2020 funded Autoship consortium aimed at demonstrating autonomy in short sea and inland waterway shipping and the European Maritime Safety Agency commissioned study into autonomous vessel safety risks and their assessment, SAFEMASS / RBAT. Meanwhile the umbrella body, Maritime UK is up to the seventh edition of its code of practice for industry players and the major ship classification societies have each published guidance documents in the intervening period. Two general purpose guidance documents are the UK's Safety-Critical Systems Club "Safety Assurance Objectives for Autonomous Systems" and the "Safety Assurance of autonomous systems in Complex Environments (SACE)" from the Assuring Autonomy International Programme. Leaning on earlier established principles and practice, management of safety risk to a tolerable level and subsequent demonstration of safety case remain pivotal to safety assurance of maritime autonomy. Functional safety is the mitigating risks of system or component failures that would otherwise cause harm. Failure and hazard analysis techniques abound and adherence to standard IEC 61508 for electrical/control systems then facilitates the assigning of Safety Integrity Levels (SILs). Good practice may be read across from safety initiatives pertaining to self-driving road vehicles. Two that stand out are mitigating for functional insufficiencies or foreseeable misuse and so called "Safety Of The Intended Functionality" (SOTIF) and standard UL4600 for overarching safety case formulation. In this paper a summary of latest thinking and development related to safety of autonomy and relevant to sea going vessels is presented.
This paper provides a benchmark of the performance of 23 classical and state-of-the-art background subtraction (BS) algorithms on visible range and near infrared range videos in the Singapore Maritime dataset. Importantly, our study indicates the limitations of the conventional performance evaluation criteria for maritime vision and proposes new performance evaluation criteria that is better suited to this problem. This paper provides insight into the specific challenges of BS in maritime vision. We identify four open challenges that plague BS methods in maritime scenario. These include spurious dynamics of water, wakes, ghost effect, and multiple detections. Poor recall and extremely poor precision of all the 23 methods, which have been otherwise successful for other challenging BS situations, allude to the need for new BS methods custom designed for maritime vision.
Abstract Uncertainty pervades the world we live in although its presence is not always acknowledged in systems engineering. Its influence on providing systems‐of‐systems defence capability is not only amplified by the aggregation of uncertainty through the systems hierarchy but also by the ambiguity in defining such capability. Worryingly, a failure to account for uncertainty can ultimately lead to a failure to satisfy the customer. Dealing with uncertainty must begin with acknowledging, understanding and synthesising the insight that literature has to offer on the topic. Such insight is summarised in this paper by reviewing the pertinent literature on the topic. A number of formalisms for expressing uncertainty exist and the pros and cons of each are surveyed. Traditional means of uncertainty analysis are also reviewed as a basis for dealing with uncertainty in defence capability. Despite the emphasis on the provision of defence capability, the topics presented in this paper are relevant to systems engineering in general.
A key enabler to monito ring the progress of aircrew training is the reliable measurement of human performance. Such reliable measurement enables a more consistent evaluation of trainee competency than can be achieved by expert observation alone. Operational training for military aircraft may involve using live, virtual and constructive experiences that are blended to simulate the real aircrew experience. Here, for example, live capability employ s real world rehearsal of exercises using operational aircraft, whilst virtual capabi lity would typically involve trainee flight of a ground based simulator and constructive capability would introduce computer piloted simulations of hostile aircraft. Despite advances in the technology supporting such training capability, the monitoring of aircrew training progress remains ultimately a subjective exercise that is open to error. Performance assessment is essentially dictated by instructor opinion and thus limited by fallacies typical of human judgment . This paper collates recent published tho ughts on aircrew performance measurement that are specific to air combat training as a source of reference for practitioners and researchers . Key contributions are a survey of sources of both objective and subjective measures of aircrew performance, and ad vertised tools for the collection of performance related raw data. Numerous sources of performance data exist in the scientific literature along with sound principles of human performa nce measurement. However, scarce reference can be found to the proven ex ploitation of such measurement such as for aircrew debrief and the tailoring of subsequent training. Specifically, l ittle eviden ce exists of validated means both to derive task performance from multiple heterogeneous measures , and to subsequent ly assess ai r combat readiness from varying performance against multiple contributing tasks. The case for further research, development and experimentation is thus suggested .
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
'/%3e%3cuse xlink:href='%23a' transform='translate(288.889 -214.211)'/%3e%3cuse xlink:href='%23a' transform='translate(108 342.749)'/%3e%3cuse xlink:href='%23a' transform='translate(469.189 342.749)'/%3e%3c/svg%3e)
