AIDANET fully closed-loop algorithm
Academic (University of Virginia / collaborators)
A neural-network based automated insulin-delivery algorithm being tested in fully closed-loop, hybrid and mixed modes. It is a research platform, not a commercial product, but directly targets the no-meal-bolus artificial-pancreas frontier.
The scorecard
Small crossover work and ongoing trials suggest feasible free-living use, but no large pivotal TIR results are available.[1]
AIDANET trials monitor safety and time below range, but evidence is still early and small.[3]
The defining feature is operation in fully closed-loop mode with no meal announcements, directly testing the core artificial-pancreas burden.[2]
Research setup uses study devices rather than a polished commercial form factor.[2]
Mean sensor glucose is a primary feasibility endpoint in FCL@Home, but broad achieved levels are not yet published.[3]
Designed to smooth dosing through adaptive networks, but variability results remain limited.[1]
At-home trials are broader than clinic-only testing, but exercise-specific performance is not yet established.[4]
Research trials compare fully closed, hybrid and mixed modes; user-facing controls are not a marketed product.[2]
Available only through research studies; no regulatory clearance or commercial path announced.[4]
Glycemic criteria are scored on the levels actually achieved in large real-world Type 1 diabetes cohorts — not the headline improvement over a trial's baseline (an improvement that looks bigger when the starting population was doing poorly). Type 2 diabetes trial data is never used to score a Type 1 system; where only improvement data exists, it informs the rationale, not the score. Freedom captures form factor and wearability, so a tubeless system is rewarded for the mobility a tubed one can't match.
The full picture
AIDANET is here because it targets the thing every commercial hybrid loop still struggles with: meals. The algorithm is being tested in fully closed-loop and hybrid modes, which makes it a useful research benchmark for how far insulin-only automation can go before insulin speed, sensing lag and missing meal context become unavoidable. It is not a product someone can choose today.
Coming soon
ETA · At-home trials recruiting/active through 2026
Sources
- [1]Miniaturized neural networks for deploying fully closed-loop automated insulin delivery · peer-reviewed · 2025-08-01
- [2]AIDANET at Home fully closed-loop vs hybrid trial (NCT07039617) · registry
- [3]Fully Closed Loop At Home feasibility trial (NCT06041971) · registry
- [4]UVA Center for Diabetes Technology current AIDANET studies · registry