Analytics & PAT
Every instrument measures something adjacent to the quantity you want; the calibration model, not the probe, is the part that fails; and a quality statement is an argument, not a reading.
Twenty-one subjects — instruments, automation and the argument that connects them. The first recurring idea: every instrument measures something adjacent to the quantity you actually want. Optical density in a well is not a cuvette measurement; trypan blue reports membrane integrity, not life; a particle count says nothing about whether anything is alive; zeta potential is a model output; a sandwich assay stops responding above a certain concentration. Reading an instrument means knowing which neighbouring quantity you are looking at.
The second: the model, not the probe, is the part that fails. PAT stands on calibration models; the step before the instrument dominates the error budget; and a fermentation with forty sensors has only a handful of real degrees of freedom — redundancy is what turns the process historian into a fault detector.
The third: a quality statement is an argument, not a reading. A passed sterility test is weak evidence about a batch, its fourteen days set by how long an injured cell takes to recover; an endotoxin unit is not a mass.
Start with bioprocess analytics and PAT; the individual instruments are best read as case studies of its calibration problem.
Articles 21
| Aseptic isolators Why a sealed box with gloves in it is a stronger sterility argument than a cleanroom — and what sets the length of its decontamination cycle. |
| Automated nucleic acid extraction Why DNA sticks to glass in salt and lets go in water — and why that equilibrium, not the robot, sets what an extraction can recover. |
| Bioprocess data historians and analytics Why a fermentation with forty sensors has only a handful of real degrees of freedom, and how that redundancy becomes a fault detector. |
| Capillary electrophoresis in biopharmaceuticals Why a molecule's charge alone separates almost nothing, and why the whole method depends on a tube narrow enough to shed its own heat. |
| Collaborative robots in bioproduction Why a robot is allowed to work beside a person without a fence — and why the same physics that makes it safe also makes it slow. |
| HPLC and UHPLC instruments Why smaller packing particles sharpen peaks, why that costs pressure, and why a detector at 280 nm cannot see a molecule with no aromatic ring in it. |
| Particle size and zeta potential analysers Why light scattering weights a population by the sixth power of diameter, and why zeta potential is a model output rather than a measurement. |
| Sterility testing systems Why a passed sterility test is a weak statement about a batch, and why the fourteen days are set by how long an injured cell takes to start growing again. |
| Automated ELISA analyzers Why a sandwich immunoassay turns colour in proportion to analyte — and why, above a certain concentration, it stops. |
| Automated sample preparation for analytics Why the step before the instrument dominates the error budget — partition chemistry, matrix effects and what automation actually removes. |
| Bioreactor gas mixing systems Why a sparger sets a mass-transfer coefficient rather than a gas flow, and why carbon dioxide, not oxygen, usually breaks the scale-up. |
| Cell counters and viability systems Why trypan blue measures membrane integrity rather than life, and why counting statistics put a floor under how well any viability number can be known. |
| Endotoxin testing systems Why a horseshoe crab clotting cascade detects lipid A, what β-glucans do to it, and why an endotoxin unit is not a mass. |
| Plate readers in bioproduction Why an optical density in a well is not a cuvette measurement, and why every detection mode trades sensitivity against what it cannot see. |
| Automated feed systems for bioreactors Why a culture is starved on purpose, and why the pump is easier to build than the signal that should be driving it. |
| Bioprocess analytics and PAT Why a process can be measured while it runs — and why the calibration model, not the probe, is the part that fails. |
| Biosafety cabinets and laminar flow hoods A HEPA filter is not a sieve, and the barrier that protects the operator is a moving air curtain nobody can see. |
| Cleanroom environmental monitoring A particle counter measures how light scatters, not what a particle is — and nothing in an ISO 14644 count says whether anything is alive. |
| Flow cytometers in bioproduction Why spectral overlap, and not the number of detectors, sets the practical ceiling on a panel. |
| Mass spectrometry in biopharmaceutical characterisation Electrospray gets a protein into the gas phase intact — and makes the signal depend on everything else eluting at that moment. |
| Robotic biobanking and sample storage Frozen is not a single state — and the damage an archive accumulates comes mostly from the moments a human opens the door. |