Concept: fully closed-loop extended-wear patch pump

This is not a product — it is a target: the insulin-delivery hardware a true artificial pancreas needs, assembled from pieces that mostly already exist but have never shipped in one box. A fully closed-loop, extended-wear patch pump would combine six properties: tubeless patch form factor; 7+ day wear; a large reservoir; ultra-fast-insulin compatibility; full phone control; and open interoperability so any algorithm — commercial or open-source — can drive it.¹²

Form factor: a patch, not a tube. The model is today's tubeless pumps — a small, waterproof Pod worn directly on the body with no tubing and no separate insulin line — but worn far longer than they currently last.¹ Tubing is the most visible burden of pump therapy; removing it is the single biggest quality-of-life upgrade a pump can offer, and this concept keeps that while extending wear.

Reservoir capacity and wear — where the gap is widest. Today's leading tubeless patch holds up to 200 units and is replaced about every 3 days (72 h).¹ Separately, the first and only infusion set cleared for up to 7 days of wear cut set changes from roughly 10 to 5 per month — but it is a tubed set, not a patch.³⁴ No tubeless pump yet reaches 7-day wear. The concept asks for both at once: a large reservoir (a developmental semi-durable patch pump already specs a 300-unit reservoir at 7-day wear) in a tubeless body.⁵

Infusion set, cannula and occlusion detection — the real unsolved problem. Extending wear is not just an adhesive challenge; it is a chemistry one. In a euglycemic-clamp study, measured insulin exposure fell about 24% from day 0 to day 7 of set wear, and other extended sets show glucose infusion rate decreasing ~28% over the wear period even when the set never alarms.⁶⁷ In the largest 7-day trial, the set's day-7 survival rate was 77.8% (about one in five failed to last the week).⁸ In a prototype set, 88% of sets were worn 7 days, yet glucose control deteriorated after day 3.⁹ These studies used different sets and methods; exposure and glucose-infusion-rate changes are not direct measurements of pump-volume loss and should not be generalized to every marketed extended-wear set. The engineering target is reliable, drift-free delivery and trustworthy occlusion detection out to day 7 — not just an adhesive that sticks.

Ultra-fast-insulin compatibility — a genuine tension. A true artificial pancreas wants the fastest possible insulin, but ultra-rapid formulations are harder on sets: in a randomized pump study, faster aspart (Fiasp) caused no confirmed occlusions but triggered premature set changes in 44% of users versus 16.7% on standard aspart.¹⁰ Designing a 7-day patch that tolerates ultra-rapid insulin is therefore a specific, named requirement — not a given.

Which CGMs, algorithms and DIY it would work with. The concept is open by design. The FDA already created the architecture: an ACE pump (alternate controller enabled) is cleared as a building block that any compatible interoperable CGM (iCGM) and any interoperable controller (iAGC) can drive.² That open-controller layer is real — Tidepool Loop was cleared in January 2023 as the first community-led innovation in diabetes to be FDA-cleared, submitted as a fully interoperable solution intended to work with compatible ACE pumps and iCGMs.¹¹ The advocacy ask is a patch pump that fully embraces this: drivable by commercial algorithms and open-source/DIY loops, with an integrated CGM, rather than locked to one ecosystem.

Phone bolusing. Full bolus and settings control from a phone is part of the target — and is already proven on a tubeless pump, so this piece is achievable rather than speculative.¹

Access and cost by region. Undefined — no such integrated product exists, so there is no price, region, or coverage to report. By analogy to other novel integrated devices, a first mover would likely launch at premium pricing before competition or insurance coverage caught up.⁵

What's coming. The pieces are converging. A developmental semi-durable patch pump with a 300-unit reservoir and 7-day wear has been previewed, alongside other tubeless multi-component designs, though analysts expect nothing before roughly 2027.⁵ Extended-wear set chemistry keeps improving across multiple companies.⁸⁷ And the interoperability scaffolding — ACE pumps, iCGMs, and a cleared open iAGC — already exists to make an open patch pump legal.²¹¹ This is an unranked design concept. Named products and clinical programmes are assessed in their own records.