The signal
between visits.
A proposed bandage that watches four wound-environment signals and shows how a bounded orchestrator could respond. Explore the entire loop with synthetic data.
A scenario, all the way through.
Choose a synthetic trace. Adjust the last sample to see the quality gate and simulated decision update.
Four-channel signal
* Wound-fluid glucose is an experimental signal shown in arbitrary units. It does not represent blood glucose. Traces use generated data, not patient measurements.
Change the signal.
Move a control to replace the final synthetic reading. Choose a scenario to reset it.
From signal to decision.
A proposed architecture, made inspectable in software.
Virtual patch
Four sensor channels sampled by a firmware-like loop. No physical board in this build.
Quality gate
Missing or implausible values interrupt the path before any proposal appears.
Trained model
A transparent classifier learns patterns from separate, generated examples.
Bounded orchestrator
Records, rechecks, or prepares a handoff. Outputs are disconnected illustrations.
The evidence travels with it.
Awaiting scenario
The handoff will summarize the generated readings and bounded decision.
Built to be
questioned.
Patchline has no physical sensors, therapy device, or patient data. A synthetic pattern score is not an infection diagnosis. The displayed 0–3 outputs are arbitrary visualization tokens and cannot be translated into a safe treatment dose.
For a real wound, clinical evaluation remains essential. This research prototype is for technical demonstration only.