Project Canary ads

Canary methaneML marks a step change in methane intelligence. methaneML is a model trained on the industry’s richest, high-frequency methane dataset and integrated with SCADA insights to ground the model in each operator’s real operating conditions. Why methaneML stands alone: ➞ Richer training data 10B+ high-frequency, high-accuracy methane measurements capturing the variability others miss – the foundation AI needs to learn accurately. ➞ Validated quantification model Physics-based, peer-reviewed, and third-party proven to mirror real methane behavior so operators know why leaks happen, how significant they are, and where to act first. ➞ Inference modeling using multi-modal data fusion methaneML fuses diverse data streams to capture both process-driven and fugitive emissions that static, deterministic models can’t. With the deepest and highest-resolution methane dataset in the industry, methaneML lays the foundation for a predictive, hardware-light future. Interested in learning more? Reach out to [email protected]

Today, we announced a portfolio of issued U.S. patents that strengthen Project Canary’s operational intelligence platform. We’ve patented our method for combining high-frequency measurements with SCADA to distinguish true methane signals from operational activity. Result: fewer false alerts, quicker investigations, and easier reporting, leading to lower overall measurement costs. 🔗 For more information, read the press release: https://lnkd.in/gWYZ4EEC

➞ Garbage In, Garbage Out: Why Validated Data Matters for Canary methaneML Machine learning is only as good as the data behind it. Canary methaneML is anchored in independently validated, peer-reviewed science – not black-box claims. Today, we announced three peer-reviewed publications that help define our foundation: • Deployment: How and where fixed sensors should be deployed to capture methane signals reliably. • Quantification: Physics-based algorithms, tested at the Methane Emissions Technology Evaluation Center (METEC), that turn those signals into accurate emission rate estimates. • Real-world complexity: Proof that our methods hold up under dynamic, real operating conditions – not just idealized test scenarios. Together, these independent publications strengthen the quantification engine behind Canary methaneML and utilize one of the industry’s most extensive high-frequency methane datasets to provide more reliable, inference-driven methane insights. When your data is validated, your model can be trusted. When it’s not, it’s just more noise. Read the release here: https://lnkd.in/gxhmR8Sq

Designed for the way utility teams actually work. Project Canary fits existing workflows instead of asking crews to change how they operate.

Survey, investigation, and reporting usually live in different systems. Project Canary brings them into one.

Leak detection that scales with your network. Learn what that looks like for gas utilities.