AID Algorithm Optimiser

This explorer uses a shared physiological engine to generate exploratory output ranges for four AID systems. The same core calculation drives all four — the adapters apply system-specific framing on top of it.

Two optimisation strategies. The four systems fall into two groups based on how they increase algorithm responsiveness:

Group 1 — Control-IQ and CamAPS FX: increase algorithm-delivered insulin. These systems track insulin on board more physiologically. The optimisation strategy is to push algorithm-delivered insulin higher — anchoring basal at up to 65% of TDD — so that only 35% remains for meal boluses. Carb ratios are set weaker to reflect this split, and the remaining bolus insulin is divided by actual carb intake. This gives the algorithm the most dynamic range to adapt to real-time glucose changes. Control-IQ achieves this through sleep mode (which tightens the target and increases basal modulation) and weaker correction factors. CamAPS FX achieves this through a lower target glucose and the same basal/bolus split. Output 1 (carb-informed) is particularly valuable for these systems — it calculates the carb ratio from the bolus fraction that remains after algorithm-delivered insulin is accounted for.

Group 2 — Medtronic 780G and Omnipod 5: lower IOB visibility via shorter AIT. These systems use a shorter active insulin time to “clear” insulin faster on paper, allowing the algorithm to deliver earlier and more frequent auto-corrections. Combined with a lower target glucose, this makes the system more responsive. The trade-off is reduced IOB visibility — the device reports less insulin on board than is physiologically active. This creates a gap that is particularly relevant during and after exercise, where hidden active insulin can contribute to hypoglycaemia.

The 1-to-5 level selector maps to these strategies per system:

*5.5 mmol/L target available in US only. In Europe, 6.1 mmol/L (110 mg/dL) is the tightest available target.

Abbreviations: CF = Correction factor rule (mmol/L). AIT = Active insulin time.

Output 1 is carbohydrate informed. Enter daily carbohydrate intake to generate it. Output 2 is rule based and always shown.

Both outputs are directional starting points. They are not prescriptions and cannot substitute for individual clinical review.

IOB trade-offs. Control-IQ uses a fixed 5h AIT — IOB is visible but decays on a linear model that diverges from physiological insulin action. CamAPS FX does not publish its IOB model, making IOB somewhat more opaque. For 780G and Omnipod 5, the shorter AIT means device-reported IOB clears faster than real insulin activity — a particular risk window during and after exercise. The IOB physiology mismatch is explored in detail in the IOB Guide. For a full comparison of how these trade-offs differ across all four systems, read The IOB Trade-Off: How Four AID Systems Make Different Bets on Insulin.