SomnetX simulates airflow and aerosol transport through the human airway, from the mouth to the deep bronchi, and tells you how much of each dose deposits in every region. Answers in days, before you commit to imaging or clinical studies.
mouthpiece
trachea
carina
bronchial tree
Pharmacopeial tests push the aerosol through a right-angle metal throat at a constant flow. For many products they underpredict what the mouth and throat catch, and so overestimate the lung dose. We run the same aerosol through a human airway with a realistic breath.
Drug mass by particle size band. No anatomy, one fixed flow.
Dose by airway region, for each flow profile and particle size you test.
Both panels are illustrative layouts, not measured values.
The EU exemption for pMDI propellants shrinks from 2027 and ends in 2030. 44% of pMDIs sold have no declared low-GWP plan.
Compare regional deposition of your current and new-propellant formulations from your own APSD data, before you commit to bridging studies.
Source: Inhalation, Dec 2025Industry estimate for a single clinical-endpoint bioequivalence study with more than 900 patients.
Regional deposition evidence to support an in vitro and modelling route, in line with FDA's work on model master files.
Source: ONdrugDelivery (Nanopharm), 2020of hospitalised COPD patients misused at least one inhaler, including breaths too weak for the device.
Test weak, typical and strong breaths in different airway geometries without recruiting a single patient.
Source: Grandmaison et al., Chronic Obstr Pulm Dis, 2024Simulation does not replace your clinical programme. It tells you which formulation, device or breathing condition deserves the expensive steps, and gives reviewers a mechanism behind your data.
Step heights indicate relative cost and effort, not exact figures.
Before we model your device, the solver is tested on the ERCOFTAC realistic-airway benchmark, where PIV measurements and published LES are available.
Mean in-plane velocity across each PIV line, normalised by the tracheal bulk velocity. Real simulation output, shown for all six stations without smoothing.
You send the airway and the question. One automated pipeline does the rest.
The airway is filled with cubes automatically. No days of manual mesh repair.
A full mouth-to-bronchi airway at 0.5 mm in about 7.4 hours, measured on our own run.
Three flow rates by four particle sizes, submitted as one batch instead of twelve projects.
Automatic CT segmentation is in development. Today we start from your STL or a segmented scan.
We resolve the turbulent laryngeal jet, track particles through it, and report deposition by anatomical region. The same model runs on your device's flow profile and your formulation's size distribution.
Lattice Boltzmann large-eddy simulation on GPUs. Sub-millimetre resolution across the whole airway, without weeks of meshing.
Millions of particles followed through the resolved flow, with drag, slip correction and gravity. Every wall impact is mapped to a region.
The same solver applied to the nasal and pharyngeal airway: flow limitation, pressure drop and collapse-prone regions in sleep-disordered breathing.
We agree the device flow profile, particle size distribution and the questions the report must answer.
→ one-page case definitionA validated reference airway, or your CT-derived airway after surface checks and region labelling.
→ labelled airway modelAirflow to statistical steady state, then particle injection across your size bins on GPU.
→ flow and particle dataRegional deposition tables, flow visualisations and a methods note you can share internally.
→ PDF report and data filesA first, bounded study to test whether simulation answers your question. Price depends on the number of flow rates, particle sizes and whether we use your airway geometry.
Send the device type, target flow rate and particle size. We reply within two working days with a scope and a fixed quote.
hello@somnetx.com