A simulation platform, purpose-built for physiological closed-loop controller design.
Ships with ready-to-use physiological models, virtual patient cohorts, optimization tools, and controller templates.
A focused environment for rapid controller iteration.
How it works
- Deploy Simulator: choose from the prebuilt library or bring your own custom physiological model.
- Iterate Controller: write Python in the browser IDE, or develop locally using your preferred tools and LLMs.
- Simulate Cohort: click "Run" in the browser or trigger via the FluxSim API to run your controller against a virtual patient cohort in the cloud.
- Analyze Outcomes: inspect per-patient loss in a heatmap, replay physiological traces, and compare performance across versions.
Prebuilt libraries.
We currently have the Schnider propofol model with a PID controller and a 200-patient virtual cohort. Our roadmap extends the simulator to include remifentanil and an MPC controller.
Models
Schnider Propofol
Minto Remifentanil
Eleveld Propofol
Controllers
PID SISO
MPC SISO
PID MIMO
MPC MIMO