Hybrid Control Chart — Integrated Statistical Process Monitoring
Hybrid Statistical Process Control Chart · Also known as: combined control chart, hybrid SPC chart, composite control chart, integrated control chart
A hybrid control chart integrates two or more classical charting schemes — most commonly a Shewhart chart with a CUSUM or EWMA chart — into a single monitoring procedure. By combining the strengths of each component, hybrid charts can detect both large, sudden shifts and small, sustained drifts in a process more effectively than any single chart alone. They are used in manufacturing quality control, healthcare monitoring, and any continuous process where rapid and sensitive detection of out-of-control conditions is critical.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use a hybrid control chart when you need to detect both sudden large shifts and small gradual drifts in the same process stream and a single chart type cannot satisfy both requirements. It is well-suited to continuous manufacturing, pharmaceutical batch processes, clinical outcome monitoring, and any setting where out-of-control conditions may appear in more than one mode. Do not use a hybrid chart when the process is only at risk of one type of shift — a single CUSUM suffices for sustained small drifts, and a Shewhart chart suffices for large jumps; adding an unnecessary component inflates complexity and can confuse operators. Also avoid hybrid charts when the sampling rate is very low or when operator training cannot support interpreting dual alarm signals correctly.
Strengths & limitations
- Combines sensitivity to small sustained drifts (CUSUM or EWMA component) with fast detection of large sudden shifts (Shewhart component), covering a wider range of out-of-control scenarios.
- Achieves shorter average run lengths to detection (ARL1) across a range of shift sizes compared to any single constituent chart.
- Flexible: the two components can be tuned independently to balance false-alarm rate and detection power for the specific process risk profile.
- Well-established in the statistical process control literature with clear design procedures and ARL tables.
- More complex to design than a single control chart; requires specifying and calibrating parameters for both components jointly to achieve the intended overall ARL0.
- Operators must understand two signaling mechanisms and interpret which component fired to diagnose the type of process disruption correctly.
- The joint false-alarm rate is higher than either component alone unless explicitly corrected during design, which can increase nuisance alarms.
- Assumes process data are approximately independent and normally distributed (or follow a known distribution); autocorrelated data or heavy tails require modified versions.
Frequently asked
Does a hybrid control chart always outperform a single chart?
Not universally. A hybrid chart achieves a shorter ARL1 across a wider range of shift sizes than either component alone, but if the process is only subject to one type of shift — say, only large sudden jumps — a well-tuned single Shewhart chart is simpler and equally effective. The hybrid pays off when the process risk profile genuinely includes both large abrupt shifts and small sustained drifts.
How do I set the parameters of a Shewhart-CUSUM hybrid?
Start by specifying the desired in-control ARL0 (commonly 370 for a false-alarm rate of about 0.27%) and the smallest shift magnitude you need to detect quickly (in units of sigma). Use published ARL tables or simulation to select the CUSUM reference value k (typically 0.5 for a 1-sigma shift) and decision interval h (typically 4–5) alongside the Shewhart limit width (typically 3 sigma). Adjust jointly until the combined ARL0 matches the target.
Should I reset the CUSUM accumulator after every signal?
Yes — after a signal is confirmed and the assignable cause is identified and corrected, reset the CUSUM accumulator (both one-sided accumulators, if used) to zero before resuming monitoring. Failure to reset means the chart carries historical memory of a corrected disturbance into future observations, inflating the likelihood of subsequent false alarms.
Can hybrid control charts be used for multivariate data?
Yes. Multivariate hybrids typically combine the Hotelling T-squared chart (analogous to the multivariate Shewhart) with a multivariate CUSUM or MEWMA chart. The design principles are the same — tune each component to complement the other — but parameter selection becomes more involved as the number of quality characteristics grows.
What if my data are autocorrelated?
Standard hybrid charts assume independence. With autocorrelated process data, fit a time-series model (e.g., ARIMA) to the observations and apply the hybrid chart to the model residuals. Alternatively, use modified CUSUM or EWMA schemes designed explicitly for autocorrelated processes. Ignoring autocorrelation inflates the false-alarm rate substantially.
Sources
- Lucas, J. M., & Crosier, R. B. (1982). Fast initial response for CUSUM quality-control schemes: Give your CUSUM a head start. Technometrics, 24(3), 199–205. DOI: 10.1080/00401706.1982.10487759 ↗
- Control chart. Wikipedia. link ↗
How to cite this page
ScholarGate. (2026, June 3). Hybrid Statistical Process Control Chart. ScholarGate. https://scholargate.app/en/experimental-design/hybrid-control-chart
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Control chartExperimental design↔ compare
- Design of experimentsExperimental design↔ compare
- Failure Mode and Effects AnalysisExperimental design↔ compare
- Process Capability AnalysisStatistics↔ compare
- Six Sigma DMAICQuality Management↔ compare
- Statistical Process ControlExperimental design↔ compare