Achieving the first successful predictive model for maritime risk management

HiLo’s industry-leading predictive model for tankers has been upgraded. The maritime industry’s safety standards are constantly improving. HiLo develops its services based on those improvements. This time, HiLo used the information provided by our team of maritime members to develop the tanker model. The sources include subscriber data, Sea web incident data and accident reports collected in 2019. Several changes have been made to provide improved safety analysis for HiLo customers.

Upgrades Include:
Segmentation of UDEs: An issue occurring in different places can have different impacts. Keeping that in mind, some of the key Undesired Events have been segmented into various categories for better granularity. Undesired Events lead to injury, ankara escort fatality and/or significant damage to assets, environment, or reputation. HiLo’s efforts have been focused on curbing these events.
Improved list of LEs and UDEs: LE, or Leading Events, are daily warning signals which are relatively harmless on their own but can lead to an Undesired Event. The list of LEs and UDEs has been expanded, adding events from the Bulker model which are also applicable to tankers. Even though many of the incidents experienced on bulker carriers and tankers are similar, the likelihood of a Leading Event turning into an Undesired Event is different for each vessel type. Hence, the events have been carefully selected for the tanker model.
Updated probabilities: HiLo members send in information regarding incidents and accidents on a monthly basis. Based on the probabilities and severities seen in those reports over the past year, improvements have been added to the model.
The new LEs and UDEs links have been added based on the information sent in from our subscribers. Analysis of HiLo data from the past year has shown cause and effect which has not been documented elsewhere. These relationships have been built into the new model. Additionally, wordings have been updated to match the bulk carrier model so that ship managers can now navigate through the multiple vessel types and compare incidents across their fleet more easily.

How Does it Work?
HiLo uses a system of advanced predictive modelling which looks at data differently from the rest of the industry.

HiLo has built a web of interconnected events, from daily warning (Leading Events) through to high impact incidents resulting in fatalities, severe injuries, or damage to assets (Undesired Events). The network is created by maritime experts and developed with HSSE professionals to ensure the events line up with existing shipping experience.
All of the cause and effect relationships are analysed to understand the likelihood of a leading event progressing to an undesired event and the likelihood that the undesired event itself results in a fatality. There are also interconnections between undesired events, such as grounding and flooding, as often one issue will cause another.
HiLo analyses thousands of event reports using this model and creates a ranking based on likelihood, severity, and frequency. Unlike analysis based on frequency alone, risk analysis is less vulnerable to fluctuation over time and focuses on areas which will make demetevler escort a tangible difference to injury and fatality rates in the future.
About HiLo
HiLo Maritime Risk Management is a not-for-profit organization that works towards saving the lives of mariners and protecting assets from shipping risks. The HiLo team includes long-serving mariners and statistics and technology experts. The company was incorporated by Shell, Lloyd’s Register Consulting and Maersk Tankers A/S with the vision to improve the levels of safety in shipping, before becoming an independent company in 2018. HiLo brings together real-world experience with a statistical model capable of predicting catastrophic maritime disasters.