Operational Vessel Data Challenge

An open modelling challenge benchmarking predictive vessel-performance models on a shared, curated dataset — to enable verifiable energy-efficiency gains in real operations.

Presented and sponsored by DNV · In collaboration with data partner Miros · Hosted by the DeWaTra COST Action

25 May 2026

Challenge opened

September 2026

Data available

31 Aug 2027

Submission deadline

Q4 2027

Winners announced
Why this challenge

Energy efficiency is shipping’s biggest near-term decarbonization lever

Maritime decarbonization cannot be achieved by clean fuels alone nor by efficiency measures alone. In the near to medium term the availability of cleaner fuels is expected to be limited, making energy efficiency the primary contributor to emission reductions — studies indicate gains of up to 16% through improved monitoring, optimization and decision support.

Achieving this at scale requires a trusted, standardized ability to measure, model and verify vessel performance under real operational and environmental conditions.

Problem 01

Trust gap

The industry lacks trusted, independently verified evidence of real savings — uptake of efficiency measures is slowed by low trust in savings claims.

Problem 02

Fragmented modelling

Different fleets use different sensor sets, preprocessing rules and modelling assumptions — results are not comparable and decision support is hard to scale.

The goal

Develop models that…

1 · Reproduce vessel performance

Model speed–power behaviour from operational + metocean data, separating environmental effects, technical condition and the contribution of technical measures.

2 · Demonstrate predictive capability

Show forecasting and generalization capability for future vessel behaviour beyond the training envelope.

The broader intent: compare modelling methodologies — physics-guided, purely data-driven and hybrid — on a common dataset, and identify key requirements for robustness and transparency in real-world performance verification.

The dataset

Several months of time-synchronized vessel data

Participants receive operational signals, environmental information and a limited set of anonymized vessel metadata (vessel type/segment and selected principal particulars).

Access control & anonymization

The dataset is shared in anonymized form for competition purposes. Distribution is controlled via registration, leading to data access on the DNV Veracity platform (free account required). Participants may not republish raw or derived data that could enable re-identification — raw or aggregated data must not be shared outside the project team. Authors are free to publish their solutions after the challenge deadline.

Timeline

From registration to award

25 May 2026

Challenge opening — registration via the form on this page

From September 2026

Registered teams receive the link to access the data on Veracity

A free Veracity account is required

October 2026 & February 2027

Online Q&A sessions with the organizers

Exact dates will be announced on this page

31 August 2027

Submission deadline — methodology, results, conceptual model, code (optional)

Submission link and results template will be published on this page

Q4 2027

Review and announcement of the winners

2028

Winning team spends one week at DNV headquarters in Norway

Evaluation criteria

How solutions are assessed

Quality of modelling

Originality of the approach

Clarity of the solution

Award

What the winning team gets

One week at DNV headquarters, Norway

Career & deployment pathways

Following the challenge, submitted solutions will be analysed collectively to extract methodological insights, with the intention to synthesize the results in a joint publication covering performance, strengths and limitations of the different approaches.

How to participate

Four steps to join

Participation is open to industrial and academic contributors affiliated with DeWaTra. New participants are invited to join via free registration on the COST Action webpage.

01

Join DeWaTra

Not a member yet? Register for free on the COST Action page.

02

Register your team

One registration per team via the registration form (one e-mail per team).

03

Access the data

Receive the link to the dataset on Veracity — a free account is required.

04

Attend the Q&A

Join the Q&A seminars in September 2026 and January 2027.

Q&A sessions — October 2026 & February 2027

Two online question & answer sessions with the organizers. The exact dates will be announced on this page — registered teams will be notified by e-mail.

Submissions

Results template

The template for submitting your results will be published here

Submissions shall include a method description and results; code may be provided optionally. Check back or register to be notified.

Resources & links

Documents and partners

Challenge presentation (PDF)

As presented at the DeWaTra conference

Registration form

Microsoft Forms — one e-mail per team

DeWaTra COST registration

Join the COST Action CA23159 for free

DNV Group R&D Maritime

Challenge organizer and sponsor

Contact

Questions about the challenge?

Contact person: Ícaro Aragao Fonseca — Group Research and Development, Maritime Programme, DNV.
Challenge authors: Oleksii Ivashenko & Ícaro Aragao Fonseca (DNV) · Gunnar Prytz & Michael H. Schmidt (Miros).

Operational Vessel Data Challenge

An open modelling challenge benchmarking predictive vessel-performance models on a shared, curated dataset — to enable verifiable energy-efficiency gains in real operations.

Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.

Strictly Necessary Cookies

Strictly Necessary Cookie should be enabled at all times so that we can save your preferences for cookie settings.

Show details
3rd Party Cookies

This website uses Google Analytics to collect anonymous information such as the number of visitors to the site, and the most popular pages.

Keeping this cookie enabled helps us to improve our website.

Show details