Multivariable closed-loop system identification
SISO and MIMO transfer functions for matrices up to 50 inputs by 50 outputs.
PITOPS is multivariable closed-loop system identification and PID tuning software from PiControl Solutions. It extracts SISO and MIMO process transfer functions directly from normal closed-loop operating data in the time domain, for matrices up to 50 inputs by 50 outputs, then uses those identified models to calculate optimal PID tuning parameters, design PLC/DCS-based advanced process control, and maintain Model Predictive Control (MPC) models — all without open-loop step tests.
Studies show that 65 to 85% of PID control loops are not fully tuned and optimized, largely because the identification methods behind proper tuning require step tests that most plants would rather avoid. PITOPS removes that barrier: it performs complete multivariable closed-loop system identification with PID loops left in Auto or Cascade mode, and with APC or MPC schemes active, so the identification work happens without disrupting production.
PITOPS identifies process models from data your historian already collects, then turns those models into optimal PID tuning, PLC/DCS-based APC parameters, and refreshed MPC models. It connects to any DCS or PLC over OPC-DA and OPC-UA, so the same identification engine works across vendors and process types. Because the same closed-loop methodology powers PID tuning, APC design, and MPC model maintenance, engineers use one tool across the full optimization workflow.
SISO and MIMO transfer functions for matrices up to 50 inputs by 50 outputs.
Identifies models directly from existing closed-loop trend data using PITOPS TFI's GRG-based 3G algorithm.
Calculates optimal controller gains from the identified process model.
Optimizes feedforward, inferential, constraint override, and other APC parameters.
Refreshes degraded Model Predictive Control models without removing the MPC from automatic mode.
Tests proposed tuning and APC changes against the identified model before touching a live loop.
Process and control engineers use PITOPS to identify closed-loop models and retune PID loops across a unit or an entire plant, without pulling loops into manual mode or scheduling a disruptive test window. It is the software behind PiControl's PID tuning consulting engagements, which typically cover 50 to 500 loops.
Advanced process control engineers use PITOPS to design PLC/DCS-based APC strategies and to extract closed-loop models that refresh degraded MPC models, without taking the underlying controller out of automatic mode. The same identification engine that tunes PID loops also feeds APC parameter design and MPC model maintenance.
For the underlying methodology, including how closed-loop identification differs from open-loop step testing, see the closed-loop system identification guide.
PITOPS's feature set is built for one outcome: an accurate process model extracted from data you already have, without disturbing the loop it describes.
Identifies SISO and MIMO transfer functions for matrices up to 50 inputs by 50 outputs.
Captures loop interactions that single-loop identification methods miss entirely.
Extracts transfer functions directly from existing trend data instead of a step test.
Removes the production disruption, scheduling delay, and off-spec risk that step tests carry.
Works while PID loops remain closed and while APC or MPC schemes stay in service.
No mode changes, no controller hand-off, and no gap in the layer already protecting the process.
Computes optimal Kc, Ti, and Td from the identified model for a chosen tuning objective.
Turns an identified model directly into deployable controller gains, not just a diagnostic.
Optimizes feedforward, inferential, and constraint-override parameters from identified models.
Runs inside the DCS or PLC you already own, with no external host computer required.
Connects to Honeywell, Emerson, Yokogawa, ABB, Siemens, Rockwell, and Schneider platforms.
One identification engine works across every DCS or PLC vendor in a multi-site plant.
Studies show that 65 to 85% of PID control loops are not fully tuned and optimized. The gap is rarely a knowledge problem, it is a testing problem. Proper system identification, the mathematical basis for correct PID tuning, PLC/DCS-based APC design, and MPC model maintenance, has traditionally required a step test, and step tests carry a cost most plants would rather not pay.
Pulling a PID controller into manual mode to run a step on the output disturbs the process directly. During the test the plant is frequently upset, off-spec, or running at reduced capacity, and operators know it.
Stepping the setpoint in Auto mode is gentler than an open-loop bump, but it is still a deliberate disturbance, usually limited to one loop at a time, and it says nothing about how that loop interacts with the loops around it.
Faced with the time, coordination, and risk a step test requires, most plants simply leave most loops untouched. That is the largest single reason so many PID loops run under-tuned for years.
PITOPS removes the structural problem: complete multivariable closed-loop identification directly from existing operating data, with PID controllers left in Auto or Cascade mode and APC or MPC schemes left active. No production disruption, and no reason to leave the other 65 to 85% of loops unaddressed.
PITOPS handles the identification and design work; the rest of the suite keeps tuning current and watches loop health afterward.
PITOPS identifies a loop and computes a tuning set once; for loops that need continuous re-tuning under changing conditions, SUPERTUNE online PID auto-tuning runs fully automatic, auto-adaptive tuning inside the DCS or PLC.
Where PITOPS designs PLC/DCS-based APC and touches up MPC models, COLUMBO focuses on refreshing full multivariable MPC models from closed-loop data without pulling the controller from automatic mode.
After PITOPS identifies and retunes a loop, APROMON real-time loop monitoring watches PID and APC loops continuously for oscillation, valve degradation, and tuning drift, flagging exactly which loops need PITOPS again next.
PITOPS identifies models on production loops across every sector PiControl serves, and the same closed-loop engine adapts to each sector's loop dynamics without a step test.
PITOPS is the closed-loop identification engine behind PiControl's engineering work: 10,000+ PID loops tuned and 50+ MPC models re-stabilized across 200+ industrial facilities since 2005, on-site and remote, worldwide. Read more in our customer success stories.
Request the PITOPS installer and licensing details for your plant.
See PITOPS identify a closed-loop model in real time, with no step test involved. A PiControl engineer will run the identification against a process relevant to your plant, and map out where PID tuning, APC design, or MPC model maintenance will pay off first.