List of datathon challenges

Published

May 24, 2026

1 Early detection, nowcasting and forecasting of animal health events

Objective:

Strengthen the capacity of WOAH and its Members to anticipate, detect and respond to emerging animal health threats by integrating early detection, nowcasting and short-term forecasting approaches. This challenge addresses issues such as delayed reporting, incomplete data and poor early warnings, which restrict situational awareness during the initial onset of outbreaks. Timely and reliable epidemic intelligence is essential for transparency, risk management and maintaining trust among members, trading partners and the international community. Improved real-time insight enables proportionate response measures and informed decision-making, thereby reinforcing WOAH’s role as a global reference for animal health information.

Data:

Data are restricted to a few priority diseases (e.g. Avian influenza, African swine fever) available from

  • WAHIS historical event notifications

  • Syndromic surveillance indicators

  • Trade, movement (FAOSTAT, UN Comtrade), environmental (ERA5, WorldClim), and contextual proxies (FAOSTAT, WorldBank, FAO HiH Geospatial data, Gridded Livestock of the World)

  • Global Forest Watch data portal – deforestation, active fire, land change indicators (API & downloads): https://www.globalforestwatch.org/ )

  • SPS notifications

  • Simulation exercises (Simex) for validation / comparison of the pipeline developed

  • Media or digital signal proxies (EIOS)

Deliverable:

An integrated analytical workflow that produces early warning signals for specific diseases and provides corrected, real-time estimates of disease dynamics and short-term forecasts of disease evolution, with the level of uncertainty being explicitly quantified.

2 AI-assisted signal extraction from event-based surveillance text, media reports, etc.

Objective:

Enhance epidemic intelligence by using artificial intelligence to extract relevant information on animal health from unstructured text sources. This challenge addresses the need to make better use of the rapidly growing informal and semi-structured information streams that can complement official surveillance systems. Strengthening event-based surveillance enables earlier identification of potential threats while respecting and integrating officially validated information. This challenge contributes to more agile horizon scanning and improved global preparedness.

Data:

  • News articles and media reports (GDELT ,RSS feeds from veterinary journals, EIOS, Social media)

  • Event-based surveillance feeds

Deliverable:

A system that can identify, classify and summarise animal health events from text sources.

3 Adjusting WOAH Monitoring Indicators using input-level indices to reduce structural bias

Objective:

Create a denominator or input‑level index representing the capacities and efforts of national Veterinary Services so that existing output indicators – disease reporting and other sanitary measures to control animal diseases or to improve animal welfare, which are considered as implementation of WOAH standards – can be normalised and compared more fairly across WOAH Membership. Scientifically, this addresses heterogeneity in data generation processes. Politically, fair and transparent comparisons are essential to avoid misinterpretation of indicators, protect trust among Members, and ensure that monitoring outputs are used constructively to support improvement rather than placing barriers to trade.

Data:

  • WAHIS disease situation and control measure data

  • Observatory indicators and monitoring outputs

  • Number of veterinary professionals extracted from IPUMS

  • WAHIS and FAOSTAT animal population data

  • World Bank Agricultural GDP and total GDP

  • World Bank income level

Deliverable:

A methodology or tool that distinguishes genuine differences in surveillance/implementation from artefacts caused by unequal human, financial, or system resources.

.

4 Trade Vulnerability and Disease Risk Along Trade and Movement Networks

Objective:

Assess how animal health events interact with trade and movement networks to generate vulnerability and disruption. Scientifically, this integrates network analysis with disease risk assessment. Politically, it informs discussions on trade resilience, proportional sanitary measures, and risk-based decision-making consistent with international standards.

Data:

  • International and regional trade data

  • Animal movement proxies and network data

  • WAHIS surveillance and outbreak data

  • Livestock population, density, and production data (FAOSTAT, Gridded Livestock of the World (GLW))

  • Geographic and administrative boundary data (WOAH internal)

  • WTO (SPS notifications)

  • Air, maritime transport (OpenSky,AIS)

  • Road density (Open street map, WB infrastructure data (?))

  • Regional (Mercosur, asean, eurostat)

Deliverable:

Indicators, analyses, or dashboards identifying vulnerable trade links and pathways associated with animal health shocks.

5 Understanding misinformation and disinformation in animal health.

Objective:

Improve understanding of the nature, drivers and dynamics of mis- and disinformation in the field of animal health in order to safeguard the effectiveness of disease prevention and control measures. From a scientific perspective, this challenge acknowledges the increasing awareness that information dynamics influence epidemiological outcomes by shaping perceptions, behaviours, and compliance with control measures. Mis- and disinformation can undermine trust in official institutions, weaken adherence to international standards and complicate crisis management during outbreaks. Generating evidence on how false or misleading narratives emerge and spread will support WOAH and its Members to communicate more proactively, transparently, and credibly.

Data:

  • Official outbreak and disease information from WAHIS

  • World Organisation for Animal Health press releases

  • Food and Agriculture Organization outbreak updates

  • World Health Organization zoonotic communications (DON)

  • Media news articles and other publicly available information sources ( GDELT

  • Media Cloud

  • RSS aggregators

  • Social Media

  • National news archives

  • Time-stamped contextual data related to major outbreak events

Deliverable:

Analytical outputs that (i) identify and classify prevalent types of mis- and disinformation in animal health and associated diseases, (ii) describe how such narratives evolve over time, particularly during high-impact outbreaks, and (iii) propose indicators or scoring frameworks to detect early signals, high-risk topics, sources, geographical areas, and diseases where mis- and disinformation is more likely to occur.

7 Forecasting Antimicrobials Intended for Use in Animals

Objective Explore the global patterns and trends in antimicrobial use (AMU) intended for use in animals using the ANIMUSE Global Database. AMU data (expressed in mg of antimicrobials adjusted by kg of animal biomass – mg/kg) will be used to identify spatial and temporal patterns across WOAH Members that have decided to be public and explore the distribution of antimicrobial classes used in food-producing animals. Country differences on antimicrobial classes, long-term trends and potential clusters of high or low AMU levels will be analysed.

Data

  • Antimicrobial use data reported by WOAH Members through the ANIMUSE System (both at Global level and for those Members decided to be public)

  • Antimicrobial quantities (kg)

  • Reporting coverage and adjusted antimicrobial quantities

  • Animal biomass denominator

  • Antimicrobial use adjusted by animal biomass (mg/kg)

  • Distribution of antimicrobial classes.

Deliverables

Participants will analyse the dataset and produce short analytical outputs addressing key questions on antimicrobial use patterns. Deliverables may include:

  • analysis of global and regional trends in AMU intensity (mg/kg)

  • cross-country comparisons of antimicrobial consumption

  • assessment of the relative contribution of antimicrobial classes to total AMU

  • identification of Members with high or low AMU intensity

  • Forecasting mg/kg by Members and WOAH regions by 2050

  • brief interpretation of observed trends and methodological considerations when analysing national AMU data.

8 AI-Assisted Extraction and Analysis of Global Veterinary Antibiotics

Objective Strengthen antimicrobial surveillance and support data verification by using artificial intelligence to extract, standardise, and consolidate information on veterinary antibiotic products from diverse, publicly available sources. The ultimate goal is to build a reliable global repository that can be used as a reference to verify national antimicrobial use (AMU) data submission.

Data

  • Official national antibiotic registration websites

  • Regional veterinary product databases

  • Pharmaceutical company product information and package inserts

  • Other reputable online sources with validated product information

Deliverables: An AI-powered system capable of automatically extracting product name, package size, active ingredient, concentration, route of administration, and target animal species and consolidating these into a structured global repository. Analytical output will highlight regional and global trends, identifying the most widely used products. For example, if a small subset of products accounts for the majority of veterinary antibiotics across countries, these can serve as priority references for data verification, reducing the need for complete global coverage of the repository.