Time Series Databases for maritime risk management -

Recently, Jesse has also been learning and expanding his knowledge, and saw an article on “Data Pre-processing of Ship Risk Management System based on TSDB” in the journal “China Ship Repair”, so this issue also wants to use this article as a basis to talk to you briefly about the application of TSDB in the field of ship risk management.

This article is only a personal opinion, please bear with meif there is any bias

TSDB to prevent ship risk management

Today, advances in technology have led to ships being better equipped and crews being more knowledgeable, but the accident rate has increased. Domino theory suggests that a casualty does not occur as an isolated event, although an injury may occur suddenly at one moment, it is the result of a series of events occurring one after another. We believe that the main factors leading to accidents are unsafe human behaviour, unsafe physical conditions and management deficiencies. It is precisely because of the high incidence of accidents that safety management of ships is a necessity. Ship safety management is the leadership, organisation and control of ship performance, technical safety, cargo safety and the safety of life and property of the people involved. In order to achieve a comprehensive and scientific risk management, the establishment of a ship risk management system is also a priority. We need to monitor in real time the relevant parameters of the important equipment and systems in the ship's cabin, analyse the relationship between changes in the relevant parameters and the failure of the ship's machinery, and identify the specific mechanisms and consequences of the risks caused.

Given that the ship risk management system is so important, what does its overall design look like? From a holistic point of view, the ship risk management system is shown in the diagram below. The ship's real-time data is transmitted to the shore terminal through the public network, where the data includes multiple parameters such as main engine load, torque, speed, fuel temperature, etc. After data pre-processing by receiving and parsing, abnormal data rejection and missing data supplementation, the data is stored in the TSDB; the data in the TSDB is subsequently read to realise the functions of real-time monitoring, over-limit alarm, data analysis and risk warning, etc.; and the corresponding data platform is built The TSDB is also used to visualise the monitoring content and subsequent analysis results. We can see that the TSDB is at the heart of the overall design.

At the same time, there are a number of issues that we need to be aware of in the above design, particularly during data transmission. For example, when the data sending device sends real-time data to the shore side, we need to ensure that the data is readable while at the same time preventing others from stealing or tampering with the data, and this needs to be done through a uniform protocol. Again, for example, data can be incorrect or missing due to problems such as network problems. Therefore, when data is received, it needs to be analysed first and then the abnormal values are removed and the missing values are filled in.Why choose TSDB as the database storage option for a ship risk management system? Generally speaking, there are two types of traditional DBMS (database management systems)

The problem with this type of database is that it has a complex structure and the data does not scale well, making it difficult to sustain large amounts of data written in real time.

The other type of database is non-relational, such as MongoDB, Redis, etc. The problem with this type of database is that it is relatively expensive to develop and maintain. In ship navigation, all kinds of data have the characteristics of high real-time, large amount of data, and all are sorted by time series. In terms of risk management system data query, it is also based on time period query, which makes the traditional database unable to meet the requirements of ship risk management system. It is not suitable for use as a database for ship risk management systems, either because of structural issues or because of development and maintenance costs. Therefore, the application of TSDB is the most suitable choice to solve the relevant problems.