Discriminating Fugitive from Process Emissions: Baseline Learning in the Field
Well Checked Systems
A field study of site-specific baseline models — how Zentinal Core™ learns normal operational emissions and isolates true fugitive anomalies.
Research
The Zensory.ai™ platform is built on patented, field-deployed research. Our findings are open and ungated — no registration wall, no sales sequence.
01 / Research program
Well Checked Systems holds patents in acoustic anomaly detection — developed on production natural gas compressors, where mechanical failures are routinely preceded by acoustic change minutes to hours before functional failure. That work underpins the Sound layer of the Zensory.ai™ platform today.
In 2026, we filed a provisional patent application covering the broader Zensory.ai™ platform: optical sensing, methane detection, and AI-driven environmental intelligence. Research continues across acoustic, computer vision, and optical gas-imaging (OGI) domains.
Our research program intersects with operations: every methodology documented here has been exercised on live wellsite equipment — not simulated data.
Research leadership
Mike Haines
Chief Technology Officer
Mike co-founded SEQTEK and brings extensive experience in artificial intelligence, machine learning, and enterprise software architecture. A former IBM systems engineer, Mike has deep technical expertise in computer vision, acoustic analysis, and sensor integration. He leads Well Checked's technology development, AI algorithm optimization, and technical infrastructure — and holds patents in acoustic anomaly detection.
02 / Publications
Papers, field studies, and technical notes from the Well Checked Systems research program.
Well Checked Systems
A field study of site-specific baseline models — how Zentinal Core™ learns normal operational emissions and isolates true fugitive anomalies.
Well Checked Systems
How Zentinal IQ™ produces audit-ready emission quantification — component-level spatial resolution, meteorological correction, and regulatory-aligned documentation.
Mike Haines · Well Checked Systems
Continuous audio capture with overlapping evaluation windows, site-specific baseline learning, and deviation scoring for predictive maintenance on reciprocating equipment.
03 / Collaboration
We work with operators, researchers, and regulators who need independent acoustic or optical gas-imaging data. If you have a site, a question, or a dataset worth examining, let's talk.
Get started
See Zensory.ai™ on your own sites. Pilot programs run alongside your existing LDAR cycles — no leap of faith required.