Cenozai Systems of Intelligence


Our platform enables your team to execute a series of workflows covering several disciplines for large volumes of data—such as a facility maintenance initiative—all in a single interface.


Powered by deep learning for computer vision and natural language understanding, our machine learning models are able to understand and extract context from your unstructured data. This enables you to combine it with your structured data to generate a comprehensive view of your digital and physical assets.


You decide whether your project requires a particular module, a workflow capsule (several modules linked together), or an entire System of Intelligence. Stack capability as your goals and process require.


Deploying IIOT infrastructure, RCM software, and other initiatives for your facilities? You might be missing a critical element. When a processed vibration signal predicts a compressor might fail, you’ll need to know where that compressor is located, what other equipment is connected to it, and what spare parts were needed in the past.

That information provides the necessary context for intelligent decision making, and it may be locked in scanned engineering drawings, purchase orders, and maintenance reports. The Facility Maintence SOI creates that essential context through a package of natural language understanding and computer vision technologies to classify engineering drawings, retrieve instrument tag data from P&IDs, fingerprint all of your documents, and more.

Asset Hierarchy Workflow

Facility Maintenance SOI: Level 1

The first step for comprehensive maintenance planning is establishing the physical and operating context of your assets. Designed according to the ISO14224:2016 framework.

Site Register Module

  • Classify your engineering drawings to easily find your most important files, such as P&IDs, single line diagrams, and safety & equipment layout drawings
  • Extract location, installation, system, consultant, fabricator, etc. to achieve levels 1-3 of ISO14224

Equipment Register Module

  • Capture key information from scanned P&IDs, including instrument and valve tags; equipment numbers; and line numbers
  • Detect subunits and link them to their corresponding equipment
  • Export a hierarchical equipment register that is compatible with engineering IM applications for hot-spotting

BOM Generation Module

  • Rapidly find parts information spread throughout hundreds of thousands of pages of vendor documents
  • Convert scanned tables to Excel

P&ID instrument tag and subunit detection


Most long-standing O&G companies have many shelved, bypassed and near field opportunities hidden in document, log, seismic and production data repositories. Uncovering them, however, typically requires lengthy processing of old typewritten documents, multidisciplinary analysis, root cause analysis, and so forth. For each discovered opportunity, it’s necessary to run a screen from then to now to check if it was subsequently exploited. Attempting to do all of this in a predominantly manual workflow may require many tedious man-years of work.

Machine-based Exploitation (MBX) uses a mix of machine learning technologies to systematically work through this plethora of analytical steps in a fraction of the time it would take a large team of human experts. All uncovered opportunities are classified as Fast, Intermediate or Slow Hydrocarbon (FISH) depending on their accessibility. MBX allows you to quickly add hundreds or thousands of barrels to your production in a fraction of the time required by conventional methods.