Tune the same process after every explanation. The simulator calculates overshoot, settling time, final error and integrated absolute error, so the exercise cannot be completed by copying a visually pleasing trend.
Follow the workflow
Begin with integral and derivative at zero, then raise proportional gain until the process responds promptly without sustained cycling. P-only control usually leaves offset because output falls as error gets smaller.
Do this in the product
Open the first-order process challenge, set Ki and Kd to zero, then compare Kp 0.5 and Kp 2.0.
Open the exerciseIntegral action accumulates error. Increase Ki gradually until final error disappears within an acceptable time. If the actuator saturates, uncontrolled integration stores a large correction and creates overshoot; the lab uses conditional integration anti-windup.
Do this in the product
Restore the suggested Kp, add Ki 0.25 and inspect final error and IAE. Save the result as a baseline project.
Open the exerciseDerivative reacts to the rate of error change and can add damping. It also magnifies measurement noise, so real controllers filter derivative and often apply it to process value rather than setpoint.
Do this in the product
Compare Kd 0 and Kd 0.05. Keep it only if the objective metrics improve.
Open the exerciseA loop tuned for a clean start may recover poorly from a load change. Acceptance should include disturbance rejection, actuator behaviour, alarms and the operating region—not just one trend capture.
Do this in the product
Pro learners can run the load-disturbance challenge and retain both graded attempts for comparison.
Open the exerciseCore concepts
The simulator creates a repeatable result; these concepts make that result transferable to real vendor software and supervised practical work.
Peak process value above setpoint, usually reported as a percentage of the requested step.
Time until the process enters an acceptance band and remains there. The band must be stated for the number to be meaningful.
Integrated absolute error adds the magnitude of error over time, penalising both large short errors and smaller persistent ones.
Continue in the workspace
Run the foundation exercise publicly, then use a subscription for advanced challenges, saved configurations, full attempt history, sharing, assigned paths and team reporting.
Related learning surfaces