Assess a gantry palletizer by converting its rated motion cycle into the cases per hour it can sustain with your actual products, pallet patterns, travel distances, and interruptions. The decisive figure is not the fastest quoted cycle, but the demonstrated output over a defined operating period, including pallet exchanges and other events that stop case placement.

A specification may look sufficient for an upstream line until the required pattern adds orientation moves, the gripper handles fewer cases than expected, or a full pallet blocks the next one. Compare automated palletizing machine specifications against the same product mix and duty assumptions, then require suppliers to show how each allowance affects delivered throughput.

Start with the production rate the palletizer must absorb

Establish the arrival rate in cases per hour at the palletizer infeed, not just the nominal output of the packaging line. Record the sustained rate, the highest rate that persists long enough to fill available accumulation, and the duration and frequency of those peaks. A machine can have adequate average capacity yet allow a queue to grow during a long production burst.

Separate scheduled production time from breaks, cleaning, and planned changeovers. Then define whether the required capacity applies during running time or across a shift. A shift-average target includes more lost time than a running-time target; comparing one supplier’s running rate with another supplier’s shift output gives a misleading result.

For mixed products, provide the case count and sequence by SKU, not only an average case size. Include the smallest and largest footprints, heights and weights, expected variation, case orientation at arrival, and the intended pallet patterns. A rate based on identical cases arranged in a favorable sequence may not represent a line that alternates products or builds several destinations at once.

Translate the cycle rating into case output

Ask exactly what the supplier calls a cycle. A complete cycle should normally cover the actions required to repeat a productive pick and placement: travel to the pick point, acquisition, movement to the pallet, release, and return or repositioning for the next pick. Determine whether the quoted time includes acceleration, deceleration, settling, gripper actuation, orientation, and any required clearance moves.

For a single repetitive cycle, the basic conversion is:

Gross cases per hour = 3,600 × cases placed per cycle ÷ seconds per complete cycle.

As an illustrative calculation, a cycle of 12 seconds placing two cases produces a theoretical 600 cases per hour. That figure assumes every cycle places two cases and repeats continuously. It does not include pallet exchange, an incomplete final pick, product starvation, or a blocked discharge. The numbers illustrate the calculation, not a typical machine rating.

Where cycles differ, calculate output over a representative pattern or sequence rather than multiplying one best-case cycle. Add the time for each distinct productive cycle and divide the total cases placed by that total time. Include partial picks and any repositioning needed between layers. If gripper preparation or upstream case collation happens in parallel, count only the portion that actually extends the critical path; do not subtract the same overlap twice.

Check where the cycle begins and ends

A time measured from pick to place omits the return unless the next cycle can begin without it. Similarly, a supplier may quote a mechanical motion time while excluding the wait for vacuum confirmation or case presentation. Request a timing diagram showing the gantry, gripper, infeed, and pallet conveyor activities on a shared timeline. It should identify which steps overlap and which must occur in sequence.

For a machine serving multiple pallet positions, request the timing for each route. The furthest pallet, a high layer requiring additional vertical clearance, or travel around an adjacent load can govern output. An average route is useful only when the expected allocation of cases among positions is also stated.

Test the payload and gripper at the required rate

Payload capacity must cover the complete moving load defined by the supplier: the product being lifted together with the gripper and any tooling or attachments included in that axis rating. Confirm whether the stated capacity applies throughout the working envelope and at the speed and acceleration used in the throughput calculation. A system may lift the load safely but require slower motion to control deflection or prevent product movement.

Gripper capacity is more than a maximum weight. Confirm the allowed case dimensions, overhang, center-of-gravity position, number of cases per pick, and spacing between cases. For vacuum tooling, the usable pick surface and packaging condition matter: seams, gaps, holes, loose film, porous material, and distorted cartons can affect reliable acquisition. Clamp tooling needs adequate access and a package that tolerates the applied pressure.

Multi-case picks can increase throughput, but only when the pattern and presentation support them. If cases must be rotated, separated, or collected before lifting, ask whether those activities occur within the quoted cycle or add time. A layer with an odd number of cases may require a smaller final pick. Mixed case dimensions can reduce the number held at once or require a tooling adjustment, even when each case is individually within the published range.

Send representative product samples or detailed packaging data for validation, including tolerances and realistic variations in fill and carton condition. Define how failed picks are detected, where rejected cases go, and whether recovery is automatic. A rate that depends on manual intervention after a missed pick needs a separate allowance for that interruption.

Account for pallet changes and shared resources

Convert gross cycling capacity into an operating estimate by adding events that occur less often than every pick. A useful calculation period is a representative production sequence containing several complete pallets, including the change from one pallet to the next. Include any slip-sheet or interlayer placement, pallet validation, full-pallet discharge, and empty-pallet presentation that applies to the proposed process.

Not every exchange necessarily stops the gantry. A system with multiple stations may continue stacking at one position while another pallet is removed. That benefit depends on controls, layout, product routing, and whether work is available for the remaining station. Ask the supplier to show the actual overlap rather than assuming that a second position eliminates exchange loss.

For a simplified time budget, use:

Estimated operating output = cases in the representative sequence ÷ total elapsed sequence time × 3,600.

The elapsed time must include productive motion and non-overlapped exchange or handling time. Apply any additional allowance for unplanned stops separately, with its definition and evidence stated. Do not multiply by a general efficiency factor if the same stoppages have already been included in the sequence. Conversely, a calculated sequence with no allowance for faults or intervention is not a shift-level commitment.

Pay particular attention to shared equipment. One pallet conveyor serving several positions, a single interlayer dispenser, or a common infeed can constrain the gantry even when its axes have spare capacity. Request the capacity and operating sequence of these components as part of the throughput calculation.

Assess accumulation, peaks, and downstream blocking

Accumulation buys time; it does not increase the palletizer’s long-term capacity. Estimate the maximum queue growth during a peak from the difference between the case arrival rate and the palletizer’s output over that interval. Compare the resulting case count with usable buffer positions, allowing for case dimensions, gaps, controls, and any zones reserved for separation or scanning.

For example, when arrivals exceed palletizing output by 120 cases per hour for 15 minutes, the queue grows by 30 cases, assuming both rates remain constant and no other losses occur. After the peak, the palletizer must have enough spare capacity to drain that queue before the next burst. A buffer sized only for one peak may fill progressively if recovery time is insufficient.

Trace the discharge path as carefully as the infeed. A full pallet that cannot leave because a wrapper, conveyor, or collection point is occupied can prevent an empty pallet from entering. Define the number of pallet positions available, exchange logic, and the maximum downstream blockage the system is expected to absorb. If an AGV collects completed pallets, include its response and clearance time in the interface assumptions rather than treating pickup as instantaneous.

Put competing specifications on the same basis

Issue each supplier the same operating data and request a throughput calculation for the same representative sequences. The comparison should distinguish values supported by a test or detailed simulation from estimates based on standard cycle times. A practical comparison records the following conditions alongside every quoted rate:

Ask for both the limiting SKU or pattern and the proposed normal mix. A supplier may meet a blended average while one demanding product falls below its required arrival rate. That shortfall can be acceptable only if the buffer and production schedule demonstrably absorb it and recover afterward.

Make the acceptance test reflect production

Before purchase, define a test sequence, duration, pass criteria, and measurement method that match the promised operating rate. Specify representative cases, pallet patterns, infeed presentation, tooling, and the number of pallet exchanges. State whether the clock runs during replenishment, fault recovery, and downstream pauses, and identify any external interruptions that will be excluded.

Record cases successfully placed on accepted pallets, elapsed time, rejected or recovered cases, and the reason for each stop. A short uninterrupted demonstration can verify motion performance, but it cannot by itself establish sustained output through recurring exchanges or product changes. Where a full production mix cannot be tested before shipment, document the untested assumptions and repeat the relevant verification after installation.

Use the results to resolve the capacity question explicitly: the selected gantry should meet the required rate under the agreed sequence, maintain a manageable queue during defined peaks, and recover before subsequent peaks arrive. Any remaining gap should be assigned to a concrete change in cycle time, buffering, pallet handling, operating schedule, or upstream rate—not concealed in an undefined efficiency percentage.