AGD Intelligence

Assemble Ellipta dry-powder inhaler — snap-fit closure and small-component seating

GSK manufactures tens of millions of Ellipta multi-dose dry-powder inhalers per year (Ware UK, Zebulon/Marietta US), assembling small molded plastic components — mouthpiece, body/chassis, dose-counter gears, foil strip and container — into a precise mechanism. Assembly requires inserting and meshing small gears, loading and indexing the medicine strip, and closing body halves via snap-fit engagement, with mouthpiece-to-container mating that must reach its operational configuration. The parts are lightweight, low-tolerance plastics where a misseated gear or incomplete snap produces a device that mis-doses or fails functional/leak testing. GSK holds patents on automated inhaler assembly and runs dedicated high-volume assembly lines (up to ~180 parts/minute on comparable industry lines), and the task is hard for a robot because correct seating and snap engagement must be felt, not merely seen. We identified this through our own research; we have not confirmed the specifics with the customer directly. This page is our researched read — a starting point for that conversation.

Readiness
stretch
Demand
strong
Source
researched
Failure tol.
medium
Tactile value
high
i

What the task is

RESEARCHED · our reconstruction

GSK manufactures tens of millions of Ellipta multi-dose dry-powder inhalers per year (Ware UK, Zebulon/Marietta US), assembling small molded plastic components — mouthpiece, body/chassis, dose-counter gears, foil strip and container — into a precise mechanism. Assembly requires inserting and meshing small gears, loading and indexing the medicine strip, and closing body halves via snap-fit engagement, with mouthpiece-to-container mating that must reach its operational configuration. The parts are lightweight, low-tolerance plastics where a misseated gear or incomplete snap produces a device that mis-doses or fails functional/leak testing. GSK holds patents on automated inhaler assembly and runs dedicated high-volume assembly lines (up to ~180 parts/minute on comparable industry lines), and the task is hard for a robot because correct seating and snap engagement must be felt, not merely seen.

To confirm with the customer

Is this the actual task and sequence? What are the real tolerances, cycle rate, and reject criteria, and which steps are today's manual bottleneck? Answering these is what turns this from a researched signal into a validated use case.