Skip to content
type1.science

Fully closed-loop, insulin-only (no meal announcement)

Multiple (Cambridge / academic / industry)

The active research frontier of insulin-only automated insulin delivery that removes carb counting and meal announcement entirely. Cambridge's CamAPS HX and other fully closed-loop systems show real gains over pump-plus-sensor in trials, but the speed of injected insulin still caps how well an unannounced meal can be handled — making faster insulins the key unlock.

Years awayEarly evidencefully-closed-loopphone-control

The scorecard

Time in range62

Fully closed-loop raised time-in-range ~13 points over pump+CGM in adults (50% vs 36%) and adolescents (45% vs 32%) — a real gain, but absolute TIR stays below the ~70% hybrid loops reach in motivated users. A July 2026 systematic review of 69 meal-announcement-free algorithm studies reports time in range spanning 65-89%, which is more encouraging — but that range is heterogeneous and much of it comes from simulation rather than from people, so the bump is deliberately small.[1]

Hypo protection75

Across trials time below 3.9 mmol/L stayed low and statistically unchanged versus comparators (~0.4-2.8%), with no severe hypoglycemia or DKA — removing meal boluses did not raise lows.

Automation level92

The defining strength: no carb counting, no meal announcement, no pre-meal bolus — the algorithm detects and corrects for meals on its own, the highest automation level demonstrated in trials.

Average glucose55

Mean glucose fell ~1.2-1.6 mmol/L vs pump+CGM in T1D trials; meaningful but capped by postprandial highs that unannounced insulin cannot fully catch.

Low variability55

Overnight control is excellent (overnight TIR up to ~96%), but daytime post-meal excursions drive most of the remaining variability.

Exercise handling45

Unannounced exercise remains a core open problem: rapid insulin-sensitivity shifts after a meal bolus are hard to predict, and reviews flag this as a leading barrier to true set-and-forget use.

Customizability60

Cambridge's adaptive algorithm self-learns continuously; the trade-off of a hands-off design is fewer manual levers, though glucose targets remain adjustable.

Access & cost32

Still investigational for fully-closed-loop T1D use: CamAPS HX is studied under research/CE-marked pathways, there is no FDA clearance for insulin-only fully-closed-loop, and no system is broadly commercially available in this mode. One commercial type 1 programme is now named — Medtronic's Vivera, built around optional meal bolusing — but it has no published type 1 outcome data yet, so this moves barely at all.[2]

Freedom (form factor)70

As a concept it assumes a modern compact or tubeless form; real-world freedom would depend on the hardware it ships on.

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

Today's best automated insulin delivery (AID) systems are hybrid closed loops: the algorithm runs basal insulin and corrections, but the person still has to announce every meal — counting carbs and bolusing — because injected insulin is too slow to catch a meal on its own.1 The research frontier described here aims to remove that last manual step entirely: a fully closed-loop, insulin-only system where you eat without telling the device anything, and the algorithm detects the meal and doses for it.1 It is not one product but a path, pursued most prominently by the University of Cambridge group (Hovorka and colleagues) alongside academic and industry teams.1

Components. Like any AID system it is a CGM + insulin pump + control algorithm. What changes is the algorithm: it adds automatic meal detection and prandial dosing so no announcement is needed. Cambridge's version is the phone-based CamAPS HX app, the fully-automated sibling of the hybrid CamAPS FX.2

Trial outcomes. In a single-center crossover trial, 26 adults with type 1 diabetes and above-target control (mean HbA1c 9.2%) used CamAPS HX with ultra-rapid lispro for 8 weeks with no meal announcement. Time in range (3.9-10.0 mmol/L / 70-180 mg/dL) rose to 50.0% versus 36.2% on their own pump plus CGM (+13.2 percentage points), mean glucose fell from 12.0 to 10.7 mmol/L, and time below 3.9 mmol/L stayed low and unchanged (0.88% vs 0.64%) with no severe hypoglycemia or ketoacidosis.2 A two-center adolescent trial (CamAPS HX with Fiasp, 8 weeks, n=24) showed the same pattern: time in range 45.2% vs 32.3% (+12.9 points), lower mean glucose, and no rise in hypoglycemia.3 In adults with type 2 diabetes the effect was even larger — time in range 66.3% vs 32.3% and HbA1c 7.3% vs 8.7% — because the slower meal dynamics of type 2 are more forgiving of insulin's lag.4

Automation level. This is the whole point: no carb counting, no meal announcement, no pre-meal bolus.23 In a supervised crossover study, UVA's fully automated RocketAP algorithm lifted time in range in the 6 hours after an unannounced meal to 83% (vs 53% with a legacy hybrid system), with no increase in hypoglycemia and improved overnight control — confirming that automatic prandial dosing can substantially blunt unannounced-meal excursions.5 Open-source AID communities pursue the same milestone using meal-detection and automatic micro-boluses.1

How good is meal detection now? (July 2026.) A systematic review published in July 2026 pooled 69 studies of meal-announcement-free algorithms (screened from 1,205 papers) and gives the clearest picture yet. Median meal-detection sensitivity was 88% and precision 93% — the algorithms usually spot the meal, and usually aren't crying wolf. But they spot it late: typical detection latency was 25-40 minutes after the meal started, by which time glucose is already climbing. Reported time in range across the studies spanned 65-89%, a wide and heterogeneous range, and the review is careful about why: much of this work is still in silico (simulated), and the simulations reported more false positives than the real-world studies did. Read it as a frontier that is genuinely maturing, not as a set of numbers you should expect on your own body today.6

The speed gap — the central limiter. The honest ceiling is postprandial: because subcutaneous insulin peaks ~1.5-2 hours after dosing, an unannounced meal spikes glucose before any reactive dose can act.7 This is why faster insulin matters more here than anywhere else, and a head-to-head trial directly tested it: faster aspart versus standard aspart in fully closed-loop with unannounced meals and exercise gave similar overall time in range (53% vs 58%, not significant), showing that today's "ultra-rapid" insulins are not yet fast enough to close the gap on their own.8 Reviews point to inhaled or other ultra-fast routes, and adjuncts, as the missing piece.1

Exercise. Unannounced activity remains a leading unsolved problem — post-meal insulin-sensitivity shifts are hard to predict, so exercise is the event most likely to break true "set-and-forget" use.1

Ages, indications, access. Trial evidence spans adolescents (13+) and adults with type 1 (and type 2) diabetes.234 CamAPS HX is CE-marked and used in research; there is no FDA clearance for insulin-only fully-closed-loop and no system is broadly commercially available in this mode — it is investigational today.12

Is industry building this? Slowly, and not all of it for us. Medtronic's next-generation Vivera algorithm is designed around optional meal bolusing — an admission that the announcement burden is the thing worth killing — and is expected as a software update to the 780G/Flex; but it comes to us via a conference roundup rather than published data, and there are no type 1 outcome numbers for it yet.9 Insulet presented a fully-closed-loop programme at ADA 2026 (the EVOLUTION 3 feasibility study, 64% time in range with no user settings) and has started the EVOLVE pivotal under an FDA IDE granted in March 2026, enrolling up to 350 adults, with a 510(k) filing targeted for 2027 and launch in 2028. Read the fine print: that programme is in insulin-treated type 2 diabetes, not type 1.10 It is a real milestone for fully closed-loop as an idea — type 2's slower meal dynamics are more forgiving of insulin's lag, which is exactly why it gets there first — but nobody should read it as a type 1 product on the way.

What's coming. Three threads converge: self-learning/adaptive algorithms that need no user tuning (a first-in-human Dexcom-based system lifted type 1 time in range from 38% to 56% with no meal announcement),11 faster insulins to shrink the postprandial gap,78 and adjunctive therapies (amylin, GLP-1, or glucagon) to blunt the spikes insulin alone can't catch.1 The adjunct thread got an interesting data point at ADA 2026: a McGill study (26 adults with type 1) reported that fully closed-loop with adjunct pramlintide (an amylin analogue) matched hybrid closed-loop with full carb counting almost exactly — about 71% vs 72% time in range, with similarly low time below range — at the cost of nausea and infusion-site discomfort. In other words, the carb-counting burden may be removable at close to zero glycemic cost if you add a drug that slows the meal down. Two caveats: this is a conference report summarized by a community source, not published data we can check, and adding pramlintide means it is no longer insulin-only.9 Together these threads define the insulin-only path to a true artificial pancreas — and the honest state of play in mid-2026 is that the algorithms have got good, and insulin is still the thing holding them back.

References

  1. Heise T, Piras de Oliveira C, Juneja R, et al. What is the value of faster acting prandial insulin? Focus on ultra rapid lispro. Diabetes Obes Metab 2022;24(9):1689-1701. doi:10.1111/dom.14773. https://pubmed.ncbi.nlm.nih.gov/35593434/ ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8

  2. Boughton CK, Hartnell S, Lakshman R, et al. Fully Closed-Loop Glucose Control Compared With Insulin Pump Therapy With Continuous Glucose Monitoring in Adults With Type 1 Diabetes and Suboptimal Glycemic Control: A Single-Center, Randomized, Crossover Study. Diabetes Care 2023;46(11):1916-1922. doi:10.2337/dc23-0728. https://pubmed.ncbi.nlm.nih.gov/37616583/ ↩ ↩2 ↩3 ↩4 ↩5

  3. Kadiyala N, Lakshman R, Allen J, et al. Fully Closed-Loop Improves Glycemic Control Compared with Pump with CGM in Adolescents with Type 1 Diabetes and HbA1c Above Target: A Two-Center, Randomized Crossover Study. Diabetes Technol Ther 2025;27(9):719-727. doi:10.1089/dia.2025.0062. https://pubmed.ncbi.nlm.nih.gov/40445776/ ↩ ↩2 ↩3

  4. Daly AB, Boughton CK, Nwokolo M, et al. Fully automated closed-loop insulin delivery in adults with type 2 diabetes: an open-label, single-center, randomized crossover trial. Nat Med 2023;29(1):203-208. doi:10.1038/s41591-022-02144-z. https://pubmed.ncbi.nlm.nih.gov/36631592/ ↩ ↩2

  5. Garcia-Tirado J, Diaz JL, Esquivel-Zuniga R, et al. Advanced Closed-Loop Control System Improves Postprandial Glycemic Control Compared With a Hybrid Closed-Loop System Following Unannounced Meal. Diabetes Care 2021;44(10):2379-2387. doi:10.2337/dc21-0932. https://pubmed.ncbi.nlm.nih.gov/34400480/ ↩

  6. Ibrahim M, Beneyto A, Contreras I, Vehi J. On the road to fully automated insulin delivery: A systematic review of meal announcement free algorithms. PLOS Digital Health 2026;5(7):e0001492. doi:10.1371/journal.pdig.0001492. https://doi.org/10.1371/journal.pdig.0001492 ↩

  7. Haahr H, Heise T. Fast-Acting Insulin Aspart: A Review of its Pharmacokinetic and Pharmacodynamic Properties and the Clinical Consequences. Clin Pharmacokinet 2020;59(2):155-172. doi:10.1007/s40262-019-00834-5. https://pubmed.ncbi.nlm.nih.gov/31667789/ ↩ ↩2

  8. Dovc K, Piona C, Yeşiltepe Mutlu G, et al. Faster Compared With Standard Insulin Aspart During Day-and-Night Fully Closed-Loop Insulin Therapy in Type 1 Diabetes: A Double-Blind Randomized Crossover Trial. Diabetes Care 2020;43(1):29-36. doi:10.2337/dc19-0895. https://pubmed.ncbi.nlm.nih.gov/31575640/ ↩ ↩2

  9. Closed-loop updates from ADA 2026 — a community conference roundup (not a manufacturer or peer-reviewed source) reporting Medtronic's Vivera algorithm (optional meal bolusing, expected as a 780G/Flex software update) and a McGill study of fully closed-loop with adjunct pramlintide in 26 adults with type 1 diabetes. https://www.diabetotech.com/blog/closed-loop-updates-from-ada-2026 ↩ ↩2

  10. "The biggest diabetes tech news out of ADA 2026." Drug Delivery Business (June 2026) — Insulet's EVOLUTION 3 fully-closed-loop feasibility study and the EVOLVE pivotal (FDA IDE granted March 2026) are in insulin-treated type 2 diabetes, with a 2028 launch targeted. https://www.drugdeliverybusiness.com/biggest-diabetes-tech-news-ada-2026/ ↩

  11. Wilkinson T, Donnelly S, Lever C, et al. First in Human Feasibility Study: Automated Insulin Delivery Utilizing a Self-Adapting Algorithm in Adults With Type 1 and Type 2 Diabetes. J Diabetes Sci Technol 2025;19322968251349528. doi:10.1177/19322968251349528. https://pubmed.ncbi.nlm.nih.gov/40607635/ ↩

Coming soon

ETA · Investigational today (no FDA clearance for insulin-only fully closed-loop; CamAPS HX CE-marked/used in research). Medtronic's Vivera is the first named commercial type 1 programme built around optional meal bolusing, but no type 1 outcome data has been published for it.

  • →Self-learning/adaptive algorithms needing no user tuning (a first-in-human Dexcom-based system lifted type 1 TIR from 38% to 56% with no meal announcement)
  • →Faster insulins (inhaled or other ultra-fast routes) to shrink the postprandial gap
  • →Adjunctive therapies (amylin, GLP-1, or glucagon) to blunt the spikes insulin alone cannot catch
  • →Medtronic's Vivera algorithm — designed around optional rather than mandatory meal bolusing; the first named commercial type 1 step in this direction, with no published type 1 outcome data yet · Software update to the 780G/Flex, reported at ADA 2026 as expected in 2027
  • →Insulet's fully closed-loop system — the EVOLUTION 3 feasibility study (64% time in range, no user settings) and the EVOLVE pivotal (FDA IDE granted March 2026, up to 350 adults) are in type 2 diabetes, not type 1; Insulet has announced no type 1 fully-closed-loop programme · 510(k) filing targeted 2027, launch 2028 — for type 2 diabetes

Sources

  1. [1]On the road to fully automated insulin delivery: a systematic review of meal announcement free algorithms · peer-reviewed · 2026-07-09 — 69 studies screened from 1,205. Median meal-detection sensitivity 88%, precision 93%, detection latency 25-40 minutes, reported time in range 65-89% — but heterogeneous, and in-silico evaluations reported more false positives than in-vivo ones.
  2. [2]Closed-loop updates from ADA 2026 · community — Community conference roundup, not a manufacturer or peer-reviewed source. Reports Medtronic's Vivera (optional meal bolusing, expected as a 780G/Flex software update) and a McGill study of fully closed-loop with adjunct pramlintide (n=26).
  3. [3]The biggest diabetes tech news out of ADA 2026 · news · 2026-06-06 — Insulet's EVOLUTION 3 feasibility study and EVOLVE pivotal are in type 2 diabetes, not type 1.