CHIRP Maritime

30/09/2026

Seafarer abandonment is not simply a contractual dispute. It can mean months without wages, expired contracts, repatriation refused, deteriorating conditions onboard, and families thousands of miles away left without the income they depend upon. In this episode of Sea Views, hosts Julia Gosling and Adam Parnell are joined by Mohamed Arrachedi, ITF FOC Network Coordinator for the Arab World and Iran at the International Transport Workers’ Federation, for an uncompromising look at one of maritime’s most persistent and disturbing problems.

https://chirp.co.uk/maritime/safety-resources/sea-views-podcast/

22/09/2026

Maintaining and improving safety depends on people trusting that they will be heard, supported and treated fairly. Speaking up about safety should not mean fearing for your job.

Splash reports that, "four in 10 seafarers either do not feel safe raising a serious concern at sea or are unsure whether they could do so without repercussions."

https://splash247.com/four-in-10-seafarers-unwilling-to-report-serious-safety-concerns/

In the article, Ellie Besley-Gould, chief executive of the Sustainable Shipping Initiative, says, “Good welfare can be delivered but those standards are not yet part of daily life for many seafarers.”

CHIRP Maritime provides an independent, confidential route to report maritime safety concerns, helping the industry learn without blame. If you have a concern, you can tell us here: https://chirp.co.uk/maritime/

More than four in 10 seafarers either do not feel safe raising a serious concern at sea or are unsure whether they could do so without repercussions, according to a new five-year review of crew welfare standards. The finding is among the starkest in Five Years into the Code of Conduct: What’s Chan...

16/09/2026

Can safety learning cross borders?

Maritime operations are international. Accident reporting systems are national - and their reports differ in structure, terminology and length.

CHIRPdb currently organises 796 UK MAIB reports and 456 US NTSB reports. During The Alan Turing Institute Data Study Group, methods developed with that material are being tested against 539 previously unseen Canadian TSBC reports.

A reduction in performance would itself be useful evidence. It would show where approaches built for one reporting environment fail to generalise and where more human review or adaptation is needed.

Learning globally means taking those differences seriously, not simply combining documents and assuming they are equivalent

Learn more about Zig Zag, who helped us develop CHIRPdb: https://www.zzag.ai/

14/09/2026

Responsible AI means preserving uncertainty. In a safety-critical sector, a confident answer is not necessarily a reliable answer.

The Data Study Group taking place this week with ZigZag Technology and The Alan Turing Institute will test extraction and classification methods against original reports and human-verified examples. It will also examine how uncertainty in automated classifications should be carried into later analysis, rather than treating every AI-generated label as fact.

This matters because investigated accidents are not a sample of every voyage and associations in reports do not automatically prove causation.

For CHIRP, responsible use of AI means traceable evidence, expert review and honest limits.

Those safeguards are essential if new analytical tools are to earn the confidence of seafarers, regulators and operators.

The DSG starts this week! Here's the original announcement. https://chirp.co.uk/news/chirp-maritime-zig-zag-technology-and-the-alan-turing-institute-to-investigate-recurring-safety-risks-in-maritime-reports/

11/09/2026

After an incident, it is easy to ask: who made the mistake? A just culture asks something more useful: what conditions made that action seem reasonable at the time?

Answering that question across hundreds of investigations is difficult. Reports are long, differently structured and distributed across jurisdictions.

Our collaboration with Zig Zag Technology and The Alan Turing Institute will test whether AI can help organise this evidence so human experts can examine recurring organisational, operational and human factors at scale.

Technology is only part of the answer. The purpose is better learning, better decisions and safer systems for the people who depend on them.

Here's Adam's blog, if you haven't seen it already: https://chirp.co.uk/news/solving-the-evidence-finding-problem-using-ai-to-uncover-safety-lessons-across-maritime-accident-reports/

10/09/2026

Recurring events rarely have a single cause

A procedure may exist but be difficult to follow. Equipment may work as designed but create an unexpected workload. Fatigue, supervision, communications and commercial pressure may combine in ways that are invisible when each factor is considered alone.

That is why systemic analysis matters.

CHIRPdb (with Zig Zag Technology) will support research into the safety factors that tend to appear together across published accident reports. The goal is to help experts identify recurring conditions and ask better questions - not to assign blame after the event.

For operators, better insight can support more focused risk controls, training and safety management.

https://chirp.co.uk/news/chirp-maritime-zig-zag-technology-and-the-alan-turing-institute-to-investigate-recurring-safety-risks-in-maritime-reports/

10/09/2026

Seafarers have long been able to make confidential safety reports using the CHIRP app. We’re pleased to share that you can additionally report directly from within the Gyro app.

With even more ways to make reports ‘there and then’ rather than waiting until you’re next at a computer you can make reports faster – which speeds up the feedback loop that improves safety for everyone – seafarers, operators and the wider industry. Take a look here: https://www.gyro.team/gyro-incident-report

Report safety concerns, unsafe conditions, bullying, fatigue and other onboard issues confidentially through GYRO app.

09/09/2026

Recommendations need visibility as well as authority.

Adam Parnell discusses how we are looking to solve the maritime sector's evidence-finding problem: https://chirp.co.uk/news/solving-the-evidence-finding-problem-using-ai-to-uncover-safety-lessons-across-maritime-accident-reports/

Maritime investigation reports contain carefully considered recommendations. The wider challenge is seeing those recommendations collectively: what has been recommended, to whom, where themes recur and what evidence exists of implementation.

Through the forthcoming Data Study Group, CHIRP, Zig Zag Technology and The Alan Turing Institute and will explore how information can be extracted from narrative reports and structured for responsible analysis.

This will not turn complex investigations into a league table or substitute automated conclusions for regulatory judgement.

It could, however, make patterns and gaps easier to examine and help policy discussions draw on a broader evidence base.

08/09/2026

Learning must travel further than one report

When a seafarer reports a concern or an investigator documents an accident, the value should extend beyond a single vessel, company or jurisdiction.

CHIRP's role has always been to help the maritime community learn without blame. Our new work with Zig Zag Technology and The Alan Turing Institute asks whether AI can help experts identify patterns across large numbers of published investigation reports; including the conditions in which people work, not simply the last action before an event.

Better analysis should ultimately mean better-informed action and safer working conditions at sea.

07/09/2026

One report can explain an accident. Thousands may reveal a system.

We are working with Zag Technology and The Alan Turing Institute to explore how data science and AI can help the maritime sector learn across accident investigations.

The upcoming Data Study Group will use CHIRPdb, which currently organises 796 UK MAIB reports and 456 US NTSB reports. A further 539 Canadian reports will test whether the methods can transfer across jurisdictions and document formats.

The aim is not to replace investigators or maritime expertise. It is to make recurring factors, safety issues and recommendations easier to identify and to turn existing knowledge into better systemic learning for everyone working at sea.

https://chirp.co.uk/news/chirp-maritime-zig-zag-technology-and-the-alan-turing-institute-to-investigate-recurring-safety-risks-in-maritime-reports/

CHIRPdb is built from publicly available investigation reports; it does not disclose confidential information submitted through CHIRP’s reporting programme.