Mixed case palletizing stacks cases of different sizes, weights, and SKUs onto one pallet, while layer palletizing places a complete tier of cases in a single motion. The two terms describe different things: layer palletizing names a placement unit, and mixed case palletizing names a SKU condition — and a robotic cell can run either mode depending on whether cases on the infeed share one height. NūMove Robotics builds a Robotic Layer Mixed Palletizer that stacks complete layers of mixed products, proving the two terms are not one axis, while Viroteq’s software instead recalculates a fresh pattern per case. DNC Automation supplies the robotic palletizing cell, conveyor, and load cell for Malaysian lines running either mode, and the load cell settles the one variable none of these systems currently measure: a case’s actual mass at the moment it reaches the robot.

Robotic Palletizing Resolves the Axis by Placement Unit, Not SKU Count

Robotic palletizing resolves the layer-versus-mixed-case axis by placement unit, not by SKU count — the keyword phrase implies a single spectrum, and the market’s own vocabulary contradicts that framing. A pallet’s SKU mix and its placement unit are two independent variables, and confusing them is the most common reading error in palletizing procurement. SKU mix asks whether a pallet carries one product or several. Placement unit asks whether the robot deposits cases one at a time or a full tier at once.

Relling Systems states plainly that a conventional inline or layer palletizer “forms and deposits whole layers at once” and reaches the highest peak rate on a single, stable SKU — tying layer formation to speed, not to product purity. NūMove’s own product line breaks the assumed pairing directly: its Robotic Layer Mixed Palletizer builds complete layers from mixed products, reaching up to 2,200 products per hour, specifically for industries with height-consistent SKUs such as beverage, beer, wine, and spirits. A robotic palletizing cell configured for layer formation can absorb mixed SKUs the instant case heights within a tier line up; it does not require single-SKU runs to justify the layer-forming tooling. The variable that actually decides the mode is height uniformity inside one tier, not the count of SKUs on the finished pallet.

That reframing changes what a specification sheet should ask for. A buyer comparing “layer palletizer” against “mixed case palletizer” quotes is really comparing two different questions bundled into one line item, and a vendor who answers only the SKU-count half of the question has left the throughput half unanswered.

What Layer Palletizing Locks: A Whole Tier, One Motion, One Height Band

Layer palletizing forms an entire tier of cases before placing it on the pallet in a single stroke, which is why it caps out at the fastest rate any of these systems publish. NūMove’s layer tool reaches 2,200 products per hour with one robot — roughly 2 to 6 times the 300-to-1,000-case range every other cited vendor publishes for single-case robotic placement. The mechanism explains the number: a layer former does not calculate, pick, and place each case individually, it assembles a full row of cases into a single carrier and transfers that carrier once. Every case inside the tier travels together, which means every case inside the tier must fit inside the same height band before the transfer motion.

One vendor’s own qualifier confirms the constraint: the layer tool “adapts really well to industries that have consistency on the heights of their SKUs,” naming beverage, beer, wine, and spirits as the fit. A tier holding a 330 ml can alongside a tall bottle of the same brand still clears this bar if both cases share a nominal height; a tier mixing shrink-wrapped multipacks with rigid cartons of different depths does not.

The layer tool also changes where gripping load lands on the case. Forks that support the bottom of the product, rather than a vacuum cup pulling from the top, avoid dropped items during the tier transfer — a detail that matters on a beverage line where the product itself, not just the case, carries most of the payload.

What Mixed Case Palletizing Locks: Per-Case Placement and a Pattern Computed Every Cycle

Mixed case palletizing places one case — or a small group — at a time, and that single design choice is what buys it height flexibility at the cost of peak rate. Viroteq’s StackrBrain calculates a new pattern for each pallet from live order data, because a fixed pallet template cannot anticipate which case arrives next when height, weight, and crush rating change case to case. XYZ Robotics names three distinct mechanism families under this same “mixed case” label: a line-to-pallet architecture that precomputes the full order pattern before cases move, a random-sequence architecture that reads incoming cases with an overhead camera and buffers them until the optimal one is available, and a pallet-to-pallet architecture where a single robot handles both depalletizing and mixed-case building without an intermediate buffer. Each family answers a different question — how much infeed control the plant already has — rather than three grades of the same technology.

Per-case placement is also what makes the pattern-stability problem solvable at all for a genuinely mixed pallet. Photoneo frames the underlying difficulty as cognitive: a human stacker must “simultaneously balance weight distribution, create interlocking patterns to prevent shifting during transit, and often sequence items according to specific store planograms” — a task that traditional fixed-pattern automation could not replicate because it has no rule for a case it has never been taught. A per-case system re-solves that placement problem on every cycle instead of executing a memorized layer, which is the only way to hold pallet stability constant while the case mix changes underneath it.

What Mixed Case Palletizing Locks: Per-Case Placement and a Pattern Computed Every Cycle

Stability Runs on Declared Case Weight — Not on a Weight the Cell Ever Measures

Every mixed case palletizing vendor on this comparison treats case weight as a fixed attribute the software already knows, not as a value the robotic cell checks. Viroteq’s placement engine “weight-balances every layer” and keeps heavy cases on lower tiers and fragile shippers on top; NūLogik’s pattern software “considers multiple stacking rules, product type, dimensions, density, weight, and crushability factor” before it ever routes a case to the robot; Photoneo’s motion-planning layer references “crush strength, weight, and destination requirements” to pick a placement location. In every one of these descriptions, weight is a master-data input — a number carried in the order file, the product catalog, or a one-time case specification — that the algorithm trusts and multiplies against a stacking rule. None of the eight sources analyzed for this comparison describes a scale, a load cell, or an in-motion weighing station anywhere on the palletizing cell itself.

That gap matters because a declared weight and a produced weight are not the same number. A case can come off an upstream filling or cartoning line light, heavy, or missing a unit, and a placement algorithm that reads the catalog value has no way to know the physical case in front of the robot disagrees with it. Photoneo’s own “Continuous Verification” step — overhead scanners re-checking the built pallet after each placement — closes the loop on geometry: it confirms a case landed at the planned position and orientation, and it flags a shift. It does not close the loop on mass, because the scan measures where the case sits, not what it weighs. A pallet can pass every geometric verification pass on Photoneo’s own cell and still carry a mis-filled case at a load-bearing tier, because the stability calculation ran on the catalog number from the start.

A load cell and indicator placed on the infeed conveyor ahead of the robot converts that catalog assumption into a measured one: the palletizing software receives the case’s actual mass at the moment it arrives, not the mass a spec sheet recorded months earlier. Nothing in the eight-source comparison changes this recommendation into a competitive claim — the gap between declared and measured weight is a documented absence across every mixed-case vendor examined, not a defect specific to one of them.

Throughput Sits on a Single Curve, and the Cases-Per-Hour Number Tells You Which Mode You’re Buying

Published throughput figures line up on one continuous curve once the placement unit is known. Manual labor runs 180 to 360 cases per hour, single-case robotic mixed palletizing runs 300 to 1,000 cases per hour, and layer-forming robotic palletizing reaches roughly 2,200 products per hour with a single robot. The gap between the bottom and the top of that curve is not “automated versus manual” — it is how many cases move per transfer motion.

Manual stacking moves one case per human motion. Single-case robotic palletizing moves one case per robot cycle, gated by a placement calculation that runs fresh each time. Layer-forming palletizing moves an entire tier per transfer, trading per-case height flexibility for a five-to-seven-fold jump in cases handled per hour.

A plant reading a “2,200 products per hour” spec sheet next to a “1,000 cases per hour” spec sheet is not comparing two robots of different quality — it is comparing two placement units, and only one of the two numbers applies once the SKU mix on a given order stops sharing a height band.

Throughput Sits on a Single Curve, and the Cases-Per-Hour Number Tells You Which Mode You’re Buying

Where Verification Actually Runs on a Mixed-Case Cell

A mixed-case palletizing cell verifies three things before it commits a placement, and none of the three is the pallet’s finished weight. It verifies case identity and dimension through 3D scanning before the pick — Photoneo’s overhead scanners capture a point cloud to confirm a case’s size, position, and orientation match what the warehouse management system expected, catching a mismatched item before the robot ever touches it. It verifies pattern feasibility through the placement software, which checks a candidate position against stacking rules — crush rating, footprint, height band — before committing the robot’s motion. It verifies placement outcome through a post-place rescan, comparing the built pallet against the theoretical plan and flagging any case that shifted during the transfer.

Each of those three checks answers “did the geometry come out as planned,” and each runs on a case attribute — dimension, expected weight, position — that entered the system as a number typed somewhere upstream. A checkweigher positioned after the cell closes the one loop none of the three checks reach: whether the pallet the robot just built actually carries the mass the order record says it should, measured on the finished load rather than assumed from the case catalog.

Summary: What Decides Layer vs Mixed Case So Far

Layer palletizing and mixed case palletizing answer two different questions, not two ends of one scale. Layer palletizing names a placement unit — a whole tier transferred in one motion, gated by height uniformity, reaching up to 2,200 products per hour. Mixed case palletizing names a SKU condition, placing one case at a time through a pattern computed fresh on every cycle, capped near 1,000 cases per hour. Both modes calculate stability from a case weight declared in an order file or product catalog; neither measures the case’s actual mass at the cell, and no vendor examined names a load cell or checkweigher anywhere on the line. The next section applies these two facts — placement unit and the measurement gap — to an actual line decision.

What Decides Layer vs Mixed Case So Far

Choosing Between Layer and Mixed Case for a Multi-SKU Malaysian Line

Choosing between layer and mixed case palletizing comes down to one question: does every case entering a single tier share a height band? A beverage or F&B co-packer running a narrow set of SKUs at consistent case heights can specify a layer-forming cell and capture the throughput ceiling — the published fit list for this equipment class names beverage, beer, wine, and spirits for exactly this reason. A distribution center or 3PL building store-ready pallets from an unpredictable SKU mix, where case height changes order to order, needs the per-case flexibility of a mixed case cell and accepts the lower peak rate that flexibility costs. Some lines need both modes on the same floor: a layer-forming cell for the high-volume, height-consistent SKUs, and a mixed case cell for the long tail of variety orders that never repeat a pattern.

Either configuration inherits the same declared-weight blind spot described above, which is why the weighing layer belongs in the specification regardless of which placement unit the line ends up buying. Malaysian plants running both filling and palletizing under one roof are positioned to close that loop directly, since the load cell that verifies a filled case’s weight before it reaches the conveyor is the same measurement discipline a checkweigher applies after the pallet leaves the robot.

The table below summarizes the decision on the four attributes this comparison actually settles — placement unit, height tolerance, peak throughput, and the mechanism family within mixed case that fits the plant’s existing infeed control.

Where the Weighing Layer Sits in the Physical Line

The weighing layer sits upstream of the robot, on the conveyor feeding the cell, not on the finished pallet, because the cell has to know a case’s real mass before the placement software decides which tier it lands on. A conveyor-mounted load cell reads the case in transit as it approaches the infeed, and the reading reaches the palletizing controller in time to override the catalog value the order file supplied. Positioning the measurement downstream — on the finished pallet only — still catches a mis-filled case, but only after the robot has already built the stack around a wrong number, which is the more expensive place to discover the mismatch. A checkweigher placed after the cell instead answers a different question: not what a single case weighs going in, but what the finished, multi-case load weighs coming out, which is the number a carrier or a receiving warehouse actually audits against the shipping manifest.

Where the Weighing Layer Sits in the Physical Line

Frequently Asked Questions

These frequently asked questions address the layer-versus-mixed-case comparison directly.

Does a layer palletizer handle more than one SKU?

Yes, provided every case entering that tier shares the same height. A robotic layer-forming palletizer builds layers from mixed products at up to 2,200 units per hour specifically for SKU families — beverage, beer, wine, spirits — that keep case height consistent across the product range.

Why is layer palletizing faster than mixed case palletizing?

A layer former transfers an entire tier of cases in one motion, while a mixed case cell places one case, or a small group, per cycle and recalculates the placement pattern each time. Published rates run roughly 2,200 products per hour for layer-forming palletizing against the 300-to-1,000-case range for single-case robotic mixed palletizing.

Does mixed case palletizing software weigh each case before placing it?

No source in this comparison describes an in-cell scale, load cell, or weighing station. Case weight enters the placement algorithm as a declared value from the order file or product catalog, which the software uses to decide tier position — it is never measured at the cell itself.

What actually gets verified after a robot places a case?

Position and orientation, through an overhead rescan comparing the built pallet to the theoretical plan. That check confirms the case landed where planned; it does not confirm the case, or the finished pallet, weighs what the order record expects.

A DNC engineer can size a layer or mixed case cell against your actual SKU height spread and specify where the load cell sits in the line — talk to our engineers to review your case data before you commit to a placement unit.