Publications and Patents
The documented foundation of Agentic Bioprocessing™
A complete scientific record. One white paper describing the architecture end-to-end. One technical note validating the perception layer on real experimental data. Five pending patents protecting the hardware and the analytics. Each document is self-contained; a reader can start anywhere.
Read the evidence, not the claims
Each document stands alone. Together they describe the foundation.
Every vendor in bioprocess analytics claims depth. Few publish the evidence. The documents below are the full case for how the platform sees, interprets, and learns biology at the single-cell level — written for technically literate readers, presented without marketing filler.
Read the white paper to understand the architecture. Read the technical note to see the evidence. Every claim elsewhere on this site rests on what these two documents defend.
If the argument is interesting, the evidence is where it lives.
Documents
Two documents. Complete. Current.
WHITE PAPER · WP01
Agentic Bioprocessing™
Begins Here.
from Hardware to Intelligence
The architectural case for single-cell label-free bioprocess intelligence. Covers the full three-layer system — from how the Dispersion Interaction Tensor is constructed, through the Biological Behaviour Interpreter that decides how to act on it, to the Federated Biological Knowledge repository that accumulates validated behavioural observations across the fleet. Includes the staged deployment path from lookup-table operation, through Shadow Mode, to autonomous control.
For Heads of Process Development, Heads of PAT, digital bioprocessing strategy leads.
TECHNICAL NOTE · TN01-R1
T-Cell Metabolic State
Classification
via the Dispersion Interaction Tensor
Full experimental validation of the Coupled Dispersion Analytics perception layer. A controlled T-cell activation experiment across three biological states — including a case that reports as a false negative under standard flow cytometry. The Dispersion Interaction Tensor resolves all three correctly from label-free impedance data alone. Complete methodology, frequency-channel construction, statistical analysis, and classification results.
For senior process scientists, analytical development leads, PAT engineers, cell therapy R&D.
The evidence base
Validated today. Expanding with partners.
The platform's architectural reach and its current experimental validation are two different things. We keep them separate here deliberately. What the hardware and the three-layer architecture are designed to support is one question. What has been tested end-to-end, under controlled conditions, and written up in a technical note is another. The map below is honest on both.
Validated today
The perception layer — Coupled Dispersion Analytics — is experimentally validated end-to-end for T-cell metabolic state classification, including the activation-with-transport-block case that conventional flow cytometry misreads. Full methodology, data, and statistical analysis are reported in TN01-R1.
In active work
CHO mAb fed-batch process monitoring — characterising the metabolic shift across a full production run under controlled conditions. A second technical note covering this dataset is in preparation.
Applications ready for validation partners
Perfusion and continuous manufacturing. NK cell and TIL cytolytic state monitoring. Scale-up and site-transfer comparability via Dispersion Interaction Tensor fingerprinting. R&D cell line and media screening at experiment throughput.
The measurement architecture is ready. What these applications need next is the process-specific validation campaign — which happens with partners, on their cells, under their conditions.
The validation portfolio is built one partner at a time.
The order is set by the partners who arrive first.
The pending patents
Five families. Two on hardware. Three on analytics.
Five pending patent families protect a single continuous architecture, from the electrode at the bottom to the fleet-level knowledge layer at the top. Two families cover the hardware — a label-free impedance cytometer that reads cellular compartments conventional instruments cannot reach, already in use in customer projects today. Three families cover the analytics that turn that measurement into Agentic Bioprocessing™. Each layer has standalone technical contribution. Together they form one integrated platform, protected at every level.
HARDWARE · TWO FAMILIES
Three-dimensional concave electrode geometry.
A proprietary electrode architecture that surrounds the sensing region in three dimensions rather than sitting flat above and below it. Enables label-free single-cell impedance measurement with the sensitivity and frequency range required to resolve intracellular compartments — not just cell size and viability. The measurement foundation of the platform and the basis of a capable cytometer in its own right.
Hydrodynamic cell positioning and separation.
A single integrated microfluidic architecture that sorts, positions, and separates cell populations within the same device. Places individual cells reproducibly in the sensing region without external sample preparation, and supports selective isolation of target populations downstream. Turns the measurement principle into a deployable instrument.
ANALYTICS · THREE FAMILIES
Coupled Dispersion Analytics — the perception layer.
The analytics that turn raw multi-frequency impedance into structured biological information. Produces calibration-free, label-free readouts of cellular state that are directly comparable across time, conditions, and sites. Removes the two largest barriers to using electrical measurements in bioprocess monitoring: calibration drift and cross-instrument variance.
Biological Behaviour Interpretation — the cognition layer.
The intelligence that reads the perception output, interprets what the cell population is doing, and translates that interpretation into process control actions. Every interpretation is validated against biophysical theory before it reaches the controller. Deployment is staged — advisory first, autonomous only after demonstrated understanding — giving customers and regulators a path to automation that does not require a leap of faith.
Federated Biological Knowledge — the evolution layer.
The knowledge architecture that accumulates validated biological observations across every deployed instrument and makes them available to new sites. New deployments inherit fleet-wide biological priors rather than starting from scratch. Privacy is inherent: contributed observations carry biological meaning but not the proprietary process recipes that produced them.
Filing numbers and claim-level detail are not reproduced on this page. The published filings are discoverable in the normal way for readers who want them.
Continue the conversation
Request the documents, or reach out directly.
Two ways forward from here. If you want the white paper and the technical note, the form takes two minutes and the documents are in your inbox the same day.
If you have a question the documents will not answer — about fit, about timing, about how this applies to your specific process — a short message is the faster route.
Follow us on LinkedIn:
InnoFluidics:
InnoFluidics Diagnostics: