Industrial AI105 R&D projectIn development · 2026
KazoilAn educational simulator of Kazakhstan's oil and gas industry with a control-system data layer
Students see oil and gas in textbooks, data engineers in other people's datasets. The goal is a simulator where the process, sensors, alarms and reporting share one model, and the data can be pulled into real industrial tools to train your own detector.
- sources in the research catalogue
- 767sources in the research catalogue
- equipment models with metrology
- 259equipment models with metrology
- terms in a three-language glossary
- 662terms in a three-language glossary
What we built
- Game design and architecture: 22 system specifications, about 7 MB of documentation
- A control-system data layer spec: tags, loops, safety logic, ISA-18.2 alarms, a historian, IEC 62443 zones
- Designed export of synthetic data: OPC UA, MQTT Sparkplug B, CSV and Parquet with labelled incidents
- A draft industry database: 259 equipment models with metrology, fields, routes, events from 1979 to 2026
- A three-language glossary of 662 terms, with the Kazakh wording under review
- A catalogue of 767 sources: every number in the simulator must have an author
Stack
- Python
- YAML
- JSON Schema
Today it is a detailed design and an industry database; the next step is a prototype of one site. It is not playable yet, and we say so.
Questions
Kazoil in brief
Who built the product, what it does and how to order something similar.
Who built Kazoil?
Kazoil is a research project of 105, a Kazakhstan IT company ("105kz" LLP, Almaty, Astana Hub resident). It is at the design stage: the 105 team is preparing the documentation, data and a prototype.
What does Kazoil do?
An educational simulator of Kazakhstan's oil and gas industry with a control-system data layer. Key features: game design and architecture: 22 system specifications, about 7 MB of documentation; A control-system data layer spec: tags, loops, safety logic, ISA-18.2 alarms, a historian, IEC 62443 zones; designed export of synthetic data: OPC UA, MQTT Sparkplug B, CSV and Parquet with labelled incidents; A draft industry database: 259 equipment models with metrology, fields, routes, events from 1979 to 2026.
Can we order a similar system for our company?
Yes. 105 builds industrial AI systems: predictive maintenance, computer vision, on-premise LLMs around a specific business: a 1–2 week problem review, a clickable prototype in 2–4 weeks, then iterative development. Source code and documentation stay with the customer. Contact: info@105.kz, +7 (771) 803-44-47.
Cases
More in this direction
Boljam
Boljam is our own predictive maintenance system: envelope-spectrum vibration diagnostics, P10/P50/P90 remaining useful life, 3D digital twins of sites, work orders and maintenance economics. The demo runs on synthetic data from a physics model; in a pilot the models are retrained on the customer's telemetry.
- Vibration diagnostics
- Remaining life P10/P50/P90
- 3D digital twins
Next step
Discuss a pilot project
Tell us about your task — we'll come back with a data audit plan and an effect estimate within two business days.
- 1A 30-minute call: the task, your current systems, who makes the decision
- 2A review of data and processes, a staged plan with time and cost ranges
- 3A 2–4 week prototype or an 8–12 week pilot with success criteria agreed upfront
NDA from the first contact
Reply within one business day