A well-placed pressure measurement can reveal more than you think.
Zhetalonians turns pressure transients into engineering information about what is happening inside a pipeline. We are an engineering decision-tool company built by people who have worked the operating problem — not a software company that discovered pipelines.
We know the cost of being wrong.
Our methods are grounded in years of technical work spanning pipeline and production systems, well intervention and coiled-tubing operations, transient-flow modelling, inverse problems, geophysics and signal interpretation, and machine learning applied to physical systems. We built Zhetalonians from the engineering problem outward, not from an AI model inward.
Founder & Technical Director
Senior scientist and inventor working at the intersection of sensor physics, inverse problems and industrial automation. Named inventor on more than a dozen granted US patents assigned to a major oilfield services company, covering:
- Non-intrusive determination of deposits, cross-sectional variation and leaks in a fluidic channel from reflected pressure waves, including machine-learning interpretation of the reflected signal
- Automatic stochastic (Bayesian / generative) inversion of physics-based pipeline models
- Coiled-tubing intervention automation: lock-up prediction, friction-reducer control, reel and stripper-pressure control from physics models and job history
- Production optimisation: gas-lift decision functions and wellbore clean-out targeting
Author or co-author of peer-reviewed work on physics-informed machine learning, large language models for industrial automation (ICSE Software Engineering in Practice, 2024) and computational acoustics for non-intrusive inspection. Identity withheld on this site for professional reasons; full credentials are provided to prospective licensees on request.
What this record means for you
The pressure-pulse inversion in this product is not a re-implementation of a textbook method: it is the method the founder patented for locating deposits and leaks from a single transducer, rebuilt as a tool you can run on your own data. The Bayesian inversion generalises the patented stochastic inversion of physics-based pipeline models. The lockup predictor descends from the patented coiled-tubing lock-up and friction-reducer work.
The same person who invented the methods wrote the validation tests, and the Validation page reports where they pass and where they do not.
Two addresses, two kinds of question
Technical
Data formats, model assumptions, whether your transient is analysable, custom forward models.
engineering@zhetalonians.comSales and licensing
Evaluation access, invoices, purchase orders, multi-seat and enterprise licensing.
inquiries@zhetalonians.com