Validating a liquid class answers a question about the past: on the day you tested it, with those tips and that liquid, it delivered the right volume. It does not tell you that the transfer you ran this morning was correct. Between the two sits everything that can drift, a clogged channel, a bad tip, a source well lower than expected, a bubble in the line, and none of it announces itself. In-line verification is the practice of checking volumes during the run itself, so a bad transfer is caught while it is happening rather than inferred from odd results a week later.
Offline validation and in-line verification answer different questions
Offline gravimetric validation characterizes the class: it establishes trueness and precision under controlled conditions and produces the numbers you trust. In-line verification is process control: it confirms that a specific transfer, in a specific run, actually happened as intended. You want both. A validated class with no in-line check can still fail silently on a bad day, and an in-line check with no validated class can only tell you that a poorly tuned transfer failed consistently. One proves the method is sound, the other proves today's run followed it.
What an in-line method can watch
The approaches fall into a few families, differing in what they measure and how much they interrupt the run.
- Pressure monitoring: the instrument watches the pressure trace inside the channel during aspiration and dispense and flags a curve that does not match the expected shape, which catches clots, empty wells, blocked tips, and broken air columns as they happen.
- Capacitive and level checks: sensing the surface confirms a well held what it should and that the tip reached liquid at all.
- Integrated volume-check kits: some platforms offer a dedicated station or consumable that measures delivered volume, gravimetrically or photometrically, as a step inside the method rather than a separate offline exercise.
- Post-run reads: a fast photometric or fluorometric read of the finished plate verifies every well after the fact, which is not strictly in-line but closes the same gap.
What it costs, and what it does not replace
In-line verification is not free. Pressure monitoring adds tuning and can raise false alarms if the thresholds are wrong for the liquid, a volume-check station adds time and hardware, and a post-run read adds a dye and a plate reader. And none of it replaces validation. A pressure trace tells you the transfer looked normal, not that it delivered exactly the target volume, so an unvalidated class with clean pressure traces is still an unvalidated class. Think of in-line checks as the seatbelt for a run whose class you already trust, not a substitute for trusting the class.
Validation tells you the class was right once. In-line verification tells you the run was right today. Regulated or not, the second question is the one your data actually depends on.