Press releases, in full.

GOTHENBURG, Sweden — 4 August 2026 — WoodIntel today launches as an independent, open platform for evidence-based wood specification, together with the publication of WoodIntel™ Scientific Methodology, Version 1.0.

The platform is designed to help architects, engineers, researchers and developers assess wood performance through transparent, traceable and reproducible information.

WoodIntel combines standards-derived parameters, species data, climate references, scientific literature and documented field observations. Its tools include service-life screening, uncertainty analysis, evidence assessment, scientific consensus mapping, contradiction detection, carbon intelligence, climate forecasting and digital-twin simulations.

Unlike conventional specification tools, WoodIntel publishes the architecture behind its results. Every modelled prediction can be linked to its methodology, inputs, assumptions, uncertainty range, data sources and component versions.

Scientific methodology published openly

The accompanying 22-page methodology report documents the platform’s complete reference architecture, including:

• data sources and provenance rules
• service-life model and calibration
• Monte Carlo uncertainty propagation
• evidence and confidence scoring
• scientific consensus and contradiction analysis
• study-quality grading
• AI models and their defined boundaries
• climate forecasting and digital-twin instruments
• open API structure
• reproducibility, limitations and versioning

The report also publishes the equations, parameter tables, uncertainty assumptions, calibration factors and the 24-case validation benchmark used by the current service-life model.

WoodIntel clearly separates three types of information: values taken from standards and datasets, values calculated by WoodIntel models, and assessments of what published scientific literature currently supports. Modelled service life is presented as a screening estimate with P10, P50 and P90 uncertainty outputs — not as a warranty or measured lifetime.

“Transparency should not stop at showing a source. It should also show how a conclusion was reached, which assumptions were made and where uncertainty remains.”

“By publishing the methodology, equations, validation results and limitations openly, we are inviting architects, engineers and researchers not only to use WoodIntel, but also to examine, reproduce and challenge it.”

AI with defined boundaries

WoodIntel uses AI for research retrieval, source-grounded question answering, classification of scientific positions and plain-language explanation.

However, language models do not generate the numerical values published by the platform. Service-life estimates, scores, carbon calculations and forecasts are produced by version-controlled deterministic models. Every AI-mediated statement is required to remain linked to retrievable evidence.

Open and independently governed

WoodIntel was initiated and funded by SiOO:X Wood Protection. The platform is brand-neutral and does not assess or recommend named commercial products.

Its methodology, evidence assessment and published conclusions are governed independently. Models and datasets are published openly so that third parties can inspect the assumptions and reproduce — or challenge — the results.

The methodology report and published WoodIntel data are released under Creative Commons CC BY 4.0. The platform’s public read API provides structured access to species data, climate zones, modelled predictions, evidence assessments, research gaps, carbon data, forecasts and approved field observations.

About WoodIntel

WoodIntel is an independent open platform for evidence-based wood specification. It connects standards, climate data, scientific literature, species information and documented field performance through transparent and reproducible models.

WoodIntel is not a certification body, accredited laboratory or research institute. It is an open platform that indexes, models and publishes evidence for timber construction.