Short answer: Vacuum grippers fail on leakage, not on lifting force. A gripper holds a load because the vacuum generator removes air faster than the cup seals leak it back; the moment leakage wins — a porous carton, a wrinkled bag, a dusty or damaged surface — the part drops, whatever the generator's specification says. Successful logistics installations are therefore designed around the worst surface in the flow, with zoned cups that isolate leaks, sensing that proves each pick, and margins sized for acceleration, not for static weight.
Generating the vacuum: pump or ejector
Two supply architectures dominate. Central vacuum pumps and blowers deliver high flow for large area grippers and multi-cup heads, with the generator kept off the moving tool. Compressed-air ejectors (venturi generators) sit at the tool: compact, fast-responding and simple to zone, at the cost of consuming compressed air — the most expensive utility in most plants. The choice follows the head's size and duty cycle, and it is an energy decision as much as a handling one: an ejector-based head running three shifts should have its air consumption in the operating-cost calculation from the start. Vacuum technology specialists such as Schmalz — whose business is exactly this component layer — publish their generator ranges in both architectures, which makes the comparison straightforward to run on real figures for a given head.
Porosity and deformation: the two enemies
Porous loads — corrugated carton, uncoated board, woven and non-woven materials — leak air through the material itself. The counters are flow, not depth of vacuum: large-area seals, foam sealing lips that conform to the surface, and generators sized to out-pump the leak. Deformable loads — bags, pouches, sacks — change shape as they are lifted, breaking seals and shifting the load's centre; they want the load supported over its whole area by a gripper matched to the footprint, rather than lifted from a few proud points. Both enemies are properties of the buyer's actual packaging, which is why gripper trials on the supplier's smooth demonstration plates prove almost nothing. The gripper families this article sits among are compared in our gripper comparison guide.
Zoning and per-pick sensing
Mixed-case work — the reality of logistics — is handled by zoning: the head is divided into independently valved sections, so only the cups over the actual case are evacuated, and a leaking zone does not collapse its neighbours. Zoning converts an assortment of case sizes from a tooling-change problem into a software selection. The companion discipline is sensing: vacuum switches per zone (or per cup on critical work) that confirm a seal before the robot accelerates, turning silent drops into detected, recoverable faults. In the cells described in our guide to vision-guided depalletising, that confirmation signal is what the vision-and-robot loop waits on before committing to a move.
Safety factors and losing power
Catalogue holding forces are static, vertical and measured on ideal surfaces. Real picks accelerate, tilt and vibrate, and standards-based design applies safety factors for exactly that — higher for horizontal shear, higher again where a dropped load endangers people, with the load's value and fragility shaping the acceptable risk. The failure case to design for explicitly is loss of supply: air or power failure mid-pick. Non-return valves and vacuum reservoirs that hold the load long enough for a controlled stop are standard practice in serious installations; a head that drops its load the instant the line hiccups is a design fault, not bad luck. Ask any supplier to walk through their head's behaviour on supply failure, with the valve and reservoir sizing shown.
When suction is the wrong answer
Some loads should not be vacuum-lifted at all: very porous or dusty surfaces that defeat any practical flow, items whose surface cannot be marked or stressed, open containers, and loads whose weight concentrates beyond a seal's area. The mature cell designs in the end-effector section of the directory typically combine principles — vacuum for the case body, clamping support for the move, or a fork under the load for the transfer — rather than forcing one physics to do every job. The buying conclusion mirrors the physics: specify the surface range and the failure behaviour first, and let the gripper be chosen against them, on your packaging, at your line speed.
Frequently asked questions
Why do vacuum grippers drop boxes that are within their rated payload?
Usually leakage, not weight. Porous carton, wrinkles, dust or tape seams let air in faster than the generator removes it, and the seal collapses. Rated payloads are measured on smooth, sealed surfaces; the buyer's worst packaging is the surface that decides real performance.
Which is better: a vacuum pump or a compressed-air ejector?
Pumps and blowers suit large heads and high duty cycles and keep energy costs down; ejectors are compact, fast and easy to zone at the tool but consume expensive compressed air. The right choice follows head size, zoning needs and duty cycle — and ejector air use belongs in the operating-cost estimate.
What happens to the load if power or air fails mid-pick?
In a properly designed head, non-return valves and a vacuum reservoir hold the load long enough for a controlled stop. A head that releases immediately on supply failure is under-engineered for logistics duty, and buyers should ask suppliers to demonstrate or document the behaviour.
When should a different gripper be used instead of vacuum?
When surfaces are too porous or dusty for any practical seal, when the load's surface cannot bear suction stress, for open containers, and where weight concentrates beyond the seal area. Combination heads — vacuum plus clamp or fork support — cover much of this middle ground.