Adverse effects of intoxicated driving have been well documented over the years, with clear conclusions. In addition, it is evident that the cognitive and neurological functions and reaction times deteriorate with the consumption of alcohol. Addressing the lack of literature on the subject, this paper focuses on studying the effects of alcohol on manoeuvring, navigational and decision-making ability in ship navigation. Ten participants – five cadets and five experienced navigation officers – volunteered and carried out a standard manoeuvre using a computer-controlled vessel simulator, under three different blood alcohol concentrations (0⋅00%, 0⋅05% and 0⋅08% BAC). Results from the simulations were used to assess the performance and the decision-making ability of participants under the influence of alcohol. In addition, the responses and behaviour of the simulated vessel when the navigators were intoxicated were analysed. Workload experienced by participants during the simulations were also assessed using the NASA Task Load Index. Findings of this preliminary study proved that the ability to make the correct decisions at the right time was drastically deteriorated when the blood alcohol concentration was increased.

The probability of a human-caused incident occurring during dynamic positioning (DP) drilling operations is determined in this paper using binary logistic regression models built with data on 42 incidents that took place during the period 2011–2015. For each case, a range of variables characterising the configuration of the DP system, weather conditions and water depth are taken into account. These variables are taken into account to develop a logistic regression model that shows the likelihood of an incident being caused by human error. The results obtained show that human-based incidents are significantly more likely to occur when there is a lower usage of thrusters. These results are useful for focusing our attention on variables that may be associated with incidents attributable to human error, as well as for setting operational limits that could help to prevent these incidents and improve safety during these operations.

Aircraft accidents are extremely rare in the aviation sector. However, their consequences can be very dramatic. One of the most important problems is runway excursions, when an aircraft exceeds the end (overrun) or the side (veer-off) of the runway. After performing exploratory analysis and hypothesis tests, a Bayesian-network-based approach was considered to provide information from risk scenarios involving landing procedures. The method was applied to a real database containing key variables related to landing operations on three runways. The objective was to analyse the effects over runway overrun excursions of failing to fulfil expert recommendations upon landing. For this purpose, the most influential variables were analysed statistically, and several scenarios were built, leading to a runway ranking based on the risk assessed.

Maritime accident statistics reveal that ship collisions are among the most frequent and severe accidents. The same statistics indicate that most of them are caused by human error, mainly due to breaches of the International Regulations for Preventing Collisions at Sea (COLREGs) and to the lack of communication between ships. There are also special situations where there is some ambiguity in the application of the COLREGs. In such occasions, and if there is no communication between the ships involved, compliance with the Rules may still end up in a collision. This article brings a new approach to Collision Avoidance Systems (CAS) and presents the earliest stages in the development of safety functions for the reduction of ship-to-ship collision risk on the high seas. These functions will help the concerned ships achieve coordinated compliance with the COLREGs. Functional safety standards are applied and, in their implementation, real, accessible electronic programmable systems (hardware and software) will be used.

This paper is the second part of a two-part research effort to find the optimal detector and estimator that minimise the integrity risk in Receiver Autonomous Integrity Monitoring (RAIM). Part 1 shows that for realistic navigation requirements, the solution separation RAIM method can approach the optimal detection region when using a least-squares estimator. This paper constitutes Part 2. It presents new methods to design Non-Least-Squares (NLS) estimators, which, in exchange for a slight increase in nominal positioning error, can substantially lower the integrity risk. A first method is formulated as a multi-dimensional minimisation problem, which directly minimises integrity risk, but can only be solved using a time-consuming iterative process. Parity space representations are then exploited to develop a computationally-efficient, near-optimal NLS-estimator-design method. Performance analyses for an example multi-constellation Advanced RAIM (ARAIM) application show that this new method enables significant integrity risk reduction in real-time implementations where computational resources are limited.

This paper describes the first of a two-part research effort to find the optimal detector and estimator that minimise the integrity risk in Receiver Autonomous Integrity Monitoring (RAIM). In this first part, a new method is established to determine a piecewise linear approximation of the optimal detection region in parity space. The paper presents examples suggesting that the optimal detection boundary lays in between that obtained using chi-squared residual-based RAIM, and that provided by Solution Separation (SS) RAIM, as one varies the alert limit requirement. In addition, these examples indicate that for realistic navigation requirements, the SS RAIM method approaches the optimal detection region. The SS RAIM detection tests will be employed in the second part of this work, which focuses on the design of non-least-squares estimators to reduce the integrity risk in exchange for a slight increase in nominal positioning error.