Models and AI tooling applied to ships, ports and environmental monitoring.
Predicts a ship's operational CO₂ efficiency grade (A to E) from its structural attributes alone, with the CO₂ fields that define the grade deliberately excluded to avoid target leakage. Gradient boosting roughly doubles the baseline accuracy on 1,570 EU MRV vessels, with deadweight and EEXI/EIV as the strongest predictors.
Pick a ship type and a real vessel from the EU MRV dataset and compare the predicted grade with the actual one. The full scikit-learn pipeline is on GitHub.
Fleet managers, maritime data scientists and researchers screening vessels for decarbonisation.
Flags likely low-efficiency ships before operational emissions data are available.
A computer-vision classifier that sorts port and marine waste into six recycling streams from a single photo. It uses a frozen CLIP ViT-B/32 model with a linear probe, reaching 94.6% cross-validated accuracy on 11,451 images. It grew out of an Omdena AI project on the Ganges river.
Upload a photo on the website, where the model runs entirely in the browser, or send it to the Telegram bot @blueportbot. Both return the category, a confidence score and disposal guidance in seconds.
Port environmental teams, dock workers and computer-vision researchers.
Puts instant, no-training waste classification in any worker's hands, with no server and no data upload.
A Model Context Protocol server that lets Claude query Brazilian port statistics (247 installations, 2010 to 2026) and estimate at-berth CO₂ with the IMO method, citing source and method in every answer.
Install the server in Claude Desktop or Claude Code and ask questions in plain language, such as cargo trends for a port or the CO₂ of a berthing.
Analysts, consultants and researchers who work with port data through AI assistants.
Turns official port statistics and a validated emissions method into answers an assistant can give and justify.
Data platforms, open datasets and pipelines built on official port and shipping records.
A research and data hub on the Brazilian port sector, updated monthly from ANTAQ. It brings together PortFlow (cargo since 2010), the Top 20 ports and terminals, Brazilian Berthings, the Port Call Monitor, Port Hubs Brazil and Europe, and a curated set of publications.
Open the hub and choose a dashboard, or browse the publications collection.
Port authorities, investors, maritime organisations and researchers needing reliable Brazilian port intelligence.
Centralises official data, analytical tools and literature in one place, replacing fragmented sources.
Annual cargo, berthings and port times for 247 Brazilian port installations, 2010 to 2026, aggregated from ANTAQ to match the official totals. Released under the ODbL licence with a Zenodo DOI.
Download the tables from Zenodo or GitHub and cite the DOI. A new version is released with each annual update.
Researchers, analysts and students who need clean, citable Brazilian port statistics.
Saves the cleaning and reconciliation work on raw ANTAQ extracts and makes results reproducible.
Ranks 3,303 shipping companies by their EU ETS carbon cost, built on the verified EU MRV 2024 dataset (12,610 ships, 87.3 Mt of ETS-scoped CO₂). At €80 per allowance the maritime bill reaches €6.99 billion a year, with the top 10 companies carrying 21.5% and MSC alone facing about €427 million.
Read the headline figures and the top-20 chart, then consult the ranking table of the top 25 Document-of-Compliance holders.
Shipping companies, maritime economists, policy analysts and researchers on EU ETS exposure.
Converts the EU MRV dataset into a company-level carbon-cost ranking: who actually pays for shipping's carbon.
IMO-based analytics for Brazilian ports: 30,972 port calls and 8,506 vessels (2019 to 2025) at Itaguaí, Itaqui and Paranaguá, with 99.1% deadweight coverage, expanding to the top 20 ports. Five modules cover the fleet profile, vessel search, port-call analytics, at-berth CO₂ with OPS potential and EU ETS exposure, and data quality. Benchmarked against the EU MRV fleet.
Browse the modules or search any vessel by name or IMO number to see its calls, size and estimated at-berth emissions. English and Portuguese.
Port authorities, terminal operators, investors and researchers in maritime decarbonisation.
The only public source of harmonised vessel-level port-call and emissions data for Brazilian ports.
A reproducible star-schema warehouse of Brazilian port operations in DuckDB, mirroring the ANTAQ open microdata, with a library of documented SQL queries for movements, berthing times, year-over-year growth, cargo mix and seasonality.
Clone the repository and run it with the bundled synthetic sample; the same schema and queries work on the real ANTAQ extracts.
Data engineers, analysts and researchers building port indicators.
Reproduces national port-sector indicators with transparent, versioned SQL instead of spreadsheets.
Emissions, traffic and sustainability evidence for ports and shipping in Brazil and Europe.
The interactive version of a peer-reviewed port sustainability framework: 84 indicators across 8 SDGs and 20 thematic dimensions (Cunha et al., 2025, Marine Policy). It won 1st place, Academic Production, at the 2026 Inova Itaqui Prize and is the framework adopted for the update of Brazil's port environmental performance index (IDA/ANTAQ).
Consult the indicator reference, then score each indicator; the overall score and performance band update automatically, in English or Portuguese.
Port sustainability managers, regulators, researchers and auditors assessing SDG performance.
A ready, peer-reviewed assessment model with instant self-scoring for reporting and gap analysis.
The first open observatory of the verified CH₄ and N₂O emissions now reported under EU MRV (2024 and 2025). On the same 11,453 ships, verified CO₂ fell 5.1% while CH₄ rose 21.2%. In 2025, 406 LNG carriers emitted 59.5% of all verified methane: decarbonisation is trading CO₂ for methane.
Read the headline findings, explore the charts by ship type and consult the full breakdown table.
Climate policy analysts, maritime researchers and journalists covering LNG and shipping decarbonisation.
Turns a 113-column regulatory dataset into the answer to one question: what does CO₂-only accounting miss?
Estimates the at-berth CO₂ that onshore power supply (OPS) would avoid for liquid bulk vessels, following the Fourth IMO GHG Study (2020), with a carbon-credit reference value. The web calculator runs on maritimeco2, an open, citable Python library.
Enter the vessel data and berthing times in the calculator, or install the library to process many calls at once.
Terminal planners, emissions analysts and researchers on maritime decarbonisation.
Standards-based estimates per port call, supporting disclosure, OPS business cases and carbon-credit assessment.
Interactive comparison of annual vessel arrivals at 15 major European ports by ship type and size class, built on official Eurostat statistics (2019 to present, closed calendar years).
Pick ports with the chips or the ranking table, filter by ship type and switch between arrivals and gross tonnage.
Port planners, maritime economists and researchers benchmarking European port profiles.
Replaces manual extraction from the Eurostat database with a one-click, always-updatable comparison.
Uses Copernicus Sentinel-5P data (2024 annual mean) to map NO₂ enhancement over 15 European ports and 2 shipping lanes. The Strait of Gibraltar lane shows +53.5% NO₂ against the surrounding sea with no city in sight, Algeciras reaches +59.6% among non-metropolitan ports, and metropolitan Piraeus climbs to +232%.
Read the enhancement chart (ports away from cities, metropolitan ports and open-sea lanes), then consult the table of port versus regional background.
Air-quality and climate researchers, port environmental teams and policy analysts.
An independent, space-based check on activity-based ship emission estimates.
Traces three decades of the "fewer ships, bigger ships" trend in European ports using Eurostat data (mar_tf_qm). Across the EU-27, average container-ship size per call rose 2.3 times (16,892 to 38,575 GT) while arrivals fell 19%, and some ports grew far more, such as Gdansk at 11.8 times.
Select ports and a ship type to compare average vessel size per call, arrivals and growth multiples, then read the thirty-year ledger.
Port planners, maritime economists and researchers studying fleet consolidation and port capacity.
Turns decades of port statistics into an interactive view of one of the sector's defining structural trends.
Estimates CO₂ from auxiliary engines at berth at the Port of Itaqui, with annual totals, monthly distribution and carbon intensity, 2022 to 2024. The method is fuel consumption multiplied by emission factor. Portuguese interface.
Browse the dashboard by year to read the totals, the monthly charts and the carbon intensity at berth.
Port operators, sustainability officers and researchers in decarbonisation.
Presents emissions and intensity ready for reporting, without building a spreadsheet model.
Free tools for literature mapping, text mining and finding collaborators.
Turns a set of PDFs into a word co-occurrence network for VOSviewer Online, detecting the most frequent terms automatically. Text is extracted entirely in the browser, with Portuguese and English stopwords removed and nothing sent to a server.
Upload the PDFs, set the network parameters (minimum frequency, co-occurrence and maximum terms) and download the JSON. Then open it at VOSviewer Online.
Researchers running bibliometric or literature reviews and text mining.
Automates co-occurrence maps from raw PDFs, a task that is slow and error-prone by hand.
Directed text mining with a controlled vocabulary. You define the keywords and the app counts their occurrences and co-occurrences across your PDFs, with grouped synonyms, compound terms, wildcards and thematic categories, then generates a JSON file for VOSviewer Online.
Upload the PDFs and list your keywords (group synonyms with "=", use "*" for variants and "[ ]" for categories, or load a domain pack). Download the JSON and open it at VOSviewer Online.
Researchers running directed bibliometric or content analysis with a predefined thesaurus.
Measures exactly the concepts the study cares about and maps them, without manual counting.
Finds the leading researchers in any field. It searches paper titles and abstracts through OpenAlex, detects the dominant topics and ranks authors by a balance of subject relevance and scholarly impact (h-index).
Type a keyword or topic and get a ranked list of experts. Suggested searches are provided as a starting point.
Researchers, PhD students and research managers scoping a field or seeking experts.
Identifies the key authors in a topic in seconds, replacing slow manual literature scanning.
Finds new potential collaborators aligned with an ORCID profile, using cosine similarity over the profile's OpenAlex topics. Existing co-authors are excluded so the results are new contacts.
Enter any ORCID iD and review the aligned researchers the tool proposes.
Researchers and groups expanding their collaboration network for proposals and studies.
Surfaces fresh, relevant collaborators automatically, saving hours of manual networking.
A municipal sustainability platform for the 27 coastal municipalities of Maranhão's Equatorial Margin, adapting the ISM-ODS method first developed for ports. It scores SDGs 7, 11, 13, 14 and 17 on a 1 to 5 maturity scale, with full data for four pilot municipalities. Portuguese interface.
Read each municipality's radar, ranking and gap to the 2030 target, then adjust indicators in the simulator to see which climate-finance calls the new score would unlock.
Municipal managers, policy analysts and researchers on local SDG progress.
Shows in real time how indicator improvements change the index, supporting evidence-based planning.