Intellect Server

Autonomous real-time engine for industrial AI, optimization, and surveillance

Intellect Server is the always-on runtime that keeps your Intellect solutions operating 24/7/365. Built as a task-oriented, message-passing system with native parallelism, it ingests live and historical plant data, executes predictive and optimization workloads, and pushes decisions back into operations—from virtual flow metering to refinery blend control—whether it’s installed on a laptop, an offshore skid, or a Kubernetes cluster 

Why operations teams deploy Intellect Server

  • Unified data plane – Pulls historian, SCADA, SQL, text/Excel, and OPC data into one CVA (Context/Variable/Alias) namespace, cleans it, and synchronizes every column in time.
  • Always-current models – Watches model accuracy online, adapts when drift is mild, and kicks off full autonomous remodels when conditions change, so estimates stay trustworthy.
  • Optimization-ready – Runs the setpoint recommendations, constraint handling, and desirability weighting that iImprove and other optimization artifacts publish, returning optimal operating points or closed-loop commands in real time.
  • System-of-systems control – Scales from a single node to multi-Intellect meshes with masters that distribute encrypted configs, route traffic, and coordinate Prometheus telemetry, so fleets of assets behave like one platform.
  • Operational assurance – Built-in self diagnostics report runtimes, configs, COM dependencies, OPC reachability, CPU/RAM, and file integrity without remoting into each box, while CARD-driven (Capacity/Availability/Reliability/DR) requirements keep infrastructure sizing honest.
  • Connector depth – Remote OPC clients run beside plant OPC servers, eliminating DCOM fragility; Intellect talks to those clients via TCP sockets so segmented control networks stay secure and data still flows.
  • Global-ready coherence – All inbound timestamps convert to ISO-8601 UTC (“2006-01-02T15:04:05.000Z”), eliminating timezone drift when wells, platforms, or factories sit on different continents.

Core engine capabilities (server-specific)

  • Data acquisition & virtualization

    • Executes scheduled or event-based pulls from PHD, PI, SQL, and file drops.
    • Uses CVA templates to normalize tag names, units, and contexts across wells, trains, or lines.
    • Aligns multi-rate test data with production history, fills gaps, and marks mandatory inputs so model packages know what to expect.

  • Model & optimization runtime

    • Hosts the model ensembles, event detectors, state machines, and optimization blocks authored in Intellect Expert/Designer/iImprove; once deployed, they are owned by the server.
    • Keeps per-model provenance (data sources, conditioning, settings), letting operators audit any prediction or setpoint.
    • Handles both recommendation-only and closed-loop writebacks to DCS/SCADA, with timers or streaming triggers controlling execution.

  • Deployment & governance

    • Single-server pilots install directly on Windows (with documented setup steps and service configs).
    • Mid-scale systems package the server and cluster services into Docker containers for repeatable rollouts; large fleets adopt Kubernetes for rolling updates and self-healing.
    • Masters plus embedded OPA enforce who can reach which Intellect instance, what variables they can see, and which admin actions they can perform.

  • Operations & reliability

    • CARD worksheets define throughput, uptime, failover, backup cadence, and disaster-recovery escalation so infrastructure, licensing, and staffing match the cost-of-downtime reality.
    • Remote OPC bridging and firewall-friendly TCP messaging mean data keeps flowing even when OT networks are segmented or DCOM hardening breaks legacy links.
    • Self-diagnostics expose configs, runtime state, and dependency status through a single websockets call—no need to remote into each node.
  •  

Value to the business

    • Compress time-to-value – Multi-source data prep, modeling, and optimization run in one place, so the gap between a multi-rate test and a production lift closes quickly.
    • Keep optimization online – Automatic drift handling, documented restart/move procedures, and multi-node failover guard against accuracy loss or downtime.
    • Governed speed – Policy-driven access (OPA) plus embedded diagnostics let teams move fast without forfeiting auditability or security.
    • Scalable trust – Whether the estate is a single pilot well or hundreds of wells, columns, and lines, Intellect Server delivers consistent, observable, and supportable performance.

Request Information