How Supervisors Use Adaptive SPC in Mining Conveyor Systems

The system maintains a rolling statistical model for every monitored parameter on every belt segment. The baseline window is configurable — typically 4 to 8 hours of recent production data — and advances continuously as new data points arrive. For each parameter, the rolling mean and rolling standard deviation are recalculated with every new reading, giving the system an accurate picture of where the process is operating right now. The rolling baseline automatically accounts for material changes, load shifts, and degradation patterns that would distort a static baseline.

UCL and LCL are recalculated at plus or minus three sigma from the rolling mean. When the process is stable and variability is low, sigma shrinks and the limits tighten — giving the supervisor earlier warning of any deviation. When a known high-variability period occurs — material change, start-up after maintenance, blend transition — sigma widens appropriately and the limits adjust to accommodate the legitimately wider distribution without generating false alarms. The result is a control chart where the limits are never wider than they need to be and never narrower than the process requires. Western Electric Rules run against these adaptive limits to detect trending patterns before individual readings breach any boundary.

Cpk is calculated continuously using the rolling mean and standard deviation against the specification range. The supervisor sees the true Cpk for every parameter on every belt segment — not a number from last week's report, but a live index that reflects the current process capability. When Cpk trends downward from 1.67 toward 1.33, the system generates an investigation alert before capability crosses the warning threshold. The Cpk/Ppk gap is also tracked: when Ppk diverges significantly from Cpk, the system flags instability that short-term analysis would miss — a pattern that typically indicates progressive degradation such as idler wear or belt cover deterioration.

No. The adaptive limit calculation is designed to do the opposite. Static limits are typically set wide enough to accommodate the full historical range of operating conditions. When the process is stable, the rolling sigma shrinks and the adaptive limits tighten — but the tighter limits reflect the actual process distribution, so the alert rate does not increase. The key difference is that static limits generate alerts for the wrong reasons (process changes that are not deviations) and miss alerts for the right reasons (gradual drift within wide boundaries). Adaptive limits eliminate the false alarms and surface the genuine events. In practice, total alert volume typically drops 50 to 70 percent, and the alerts that remain are the ones the supervisor needs to act on. Book a Demo to see comparative alert rate data from static versus adaptive SPC on mining conveyor parameters.

The adaptive SPC platform includes configurable transition window management. When a known process change is registered — material type switch, belt start-up after maintenance, speed change for production rate adjustment — the system enters a transition window where the rolling baseline is calculated on a shorter data window (typically 30 to 60 minutes instead of the standard 4 to 8 hours). This allows the limits to converge to the new operating conditions rapidly without generating false alarms during the convergence period. The transition window duration and the shortened baseline period are configurable per conveyor segment and per transition type. When the process stabilises at the new operating point, the standard baseline window resumes and the limits tighten to reflect the stable process. The supervisor sees the transition status on the control chart and knows that the system is in a learning phase. Talk to an expert about configuring transition windows for your conveyor network's operating profile.

Yes. All eight Western Electric Rules run against the adaptive limits, not static boundaries. The rule thresholds — Zone A (plus or minus 2 sigma), Zone B (plus or minus 1 sigma), and the limit itself (plus or minus 3 sigma) — are recalculated automatically every time the rolling baseline updates. This means the rules detect trend patterns relative to the current process distribution, not relative to a static normal that may no longer apply. A rule violation in the adaptive context means the current trend is unusual for the process as it is running right now — not unusual for the process as it was running last quarter. The practical effect is that Western Electric Rules become dramatically more useful. Static Western Electric Rules on a conveyor motor current chart generate frequent false alarms because the process mean shifts with load and material changes. Adaptive Western Electric Rules only fire when the trend is genuinely unusual relative to the current baseline. Book a Demo to see Western Electric Rules operating against adaptive limits on live conveyor parameter data.

Yes. ISO 9001 Clause 7.5 requires that documented information be controlled and maintained. For control limits, this means every limit change must have a documented rationale. iFactory's adaptive SPC platform addresses this through an automatic limit change log that records every recalculation — the timestamp, the triggering event (baseline shift, material change, transition window), the previous UCL and LCL values, the new values, and the statistical basis for the recalculation (the data window and sigma calculation used). The argument for adaptive limits in an ISO 9001 context is straightforward: limits calibrated on process data that is 6 to 12 months old are less defensible than limits that are recalculated continuously against current process data, with every recalculation logged and traceable. An auditor reviewing an adaptive limit history sees a controlled, documented process with statistically justified adjustments. A static limit history shows limits that were set once and never reviewed until the next capability study — which is harder to defend as an active quality management practice. Talk to an expert about configuring the adaptive limit change log format for your QMS documentation requirements.