A batch hold is the controlled way a running phase is paused into a defined safe condition so that a batch can weather an upset without being lost. Restart is the equally controlled way it resumes, from a known point, once the problem is cleared. Together they are the backbone of real-world batch exception handling, because they let a batch survive interruptions that would otherwise force a scrap. The subtlety is that a proper restart is not simply re-running the phase; it is dedicated logic that brings the batch back on track without double-charging or thermal shock.
batch hold and restart in one line: A batch hold pauses a phase into a defined safe hold state and waits, rather than terminating the batch. A restart resumes the held phase from a known restart point using dedicated restart logic. Holds can be triggered automatically by monitored conditions or commanded by an operator, and restart logic ensures the batch resumes safely rather than blindly repeating the phase.
When a hold is triggered, the phase does not just freeze in place. It runs its holding logic to bring the equipment into a defined safe condition before settling into the held state. For a charge phase that might mean closing the feed valve and stopping the pump; for a heating phase it might mean holding the current temperature or backing off the steam. The point is that a hold moves the equipment to a deliberately chosen safe state, not wherever it happened to be.
Once in the held state, the phase waits indefinitely. Material is secured, the equipment is stable, and no time-critical action is left dangling. This is why a hold buys time: an operator can investigate an alarm, wait for a utility to recover, or resolve a quality question while the batch sits safely. A well-designed hold state is one the process can tolerate for a while without degrading the product.
A hold is distinct from a pause. A pause typically stops the phase at a convenient logical break point and holds progress there, a lighter action suited to brief interruptions. A hold, by contrast, actively drives the equipment to a defined safe state and is meant for genuine exceptions. Choosing between them depends on how serious the interruption is and how the process behaves if it is interrupted mid-action.
Holds arise in two ways. An automatic hold is triggered by a monitored condition: a deviation exceeds a limit, an interlock trips, a utility fails, or an expression evaluates true. The exception logic issues the hold command without waiting for a human, which is essential when the process would be damaged by any delay. Automatic holds are how a batch protects itself faster than an operator could react.
An operator-commanded hold is issued deliberately by a person, for reasons the automation cannot see. An operator might hold a batch because an upstream unit is running late, because a sample needs to be pulled, or because a supervisor wants to check something before the batch proceeds. The hold mechanism is identical; only the trigger differs. Both land the phase in the same held state and both are captured in the batch record.
Whatever the trigger, the hold state gives a consistent, safe place to stand. Because automatic and commanded holds share the same held state and the same restart path, the rest of the system does not need to care why the hold occurred. This uniformity is what lets operators respond to any hold with the same procedure and lets the batch record log every hold in the same structured way.
The hardest part of hold and restart is the restart, and the reason is that you usually cannot just re-run the phase from the top. Consider a charge phase held after adding 600 kg of a 1000 kg target. Re-running from zero would add another 1000 kg on top, double-charging the batch and ruining it. Correct restart logic must recognize that 600 kg is already in and resume the charge for only the remaining 400 kg.
Thermal and timing effects raise the same issue. A reaction held mid-heat should resume from its current temperature, not restart the ramp as if the vessel were cold, or it may overshoot or shock the contents. A timed hold should account for the time already elapsed. Restart logic exists precisely to reconcile the phase's intended endpoint with the actual state the equipment reached before the hold, so the batch converges on the right result.
For SCADA and cloud monitoring, well-defined restart points are also what make remote intervention safe. When each hold captures the restart point and the batch state, an operator monitoring a distributed operation can hold a phase, see exactly where it stopped, and resume it with confidence that the system will pick up correctly rather than repeat completed work. Restart logic, not brute re-execution, is what turns a hold from a last resort into a routine, low-cost recovery.
A hold actively drives the equipment to a defined safe state and settles into the held state, meant for genuine exceptions. A pause typically stops the phase at a convenient logical break point and holds progress there, a lighter action for brief interruptions. Which to use depends on how serious the interruption is and how the process tolerates being stopped mid-action.
Because the phase may have already done part of its work. A charge phase held after adding 600 of 1000 kg would double-charge if re-run from zero. Restart logic resumes from a known restart point, adding only the remaining amount or continuing a heat from the current temperature, so the batch reaches the right endpoint without repeating completed actions.
An automatic hold is triggered by a monitored condition without human intervention: a deviation exceeding a limit, an interlock tripping, a utility failing, or an exception expression evaluating true. Automatic holds protect the batch faster than an operator could react, and they land the phase in the same held state as an operator-commanded hold, following the same restart path.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.