A pallet that contains several SKUs is rarely difficult because of the robot itself. The challenge is making sure each case arrives in the right order, can be identified correctly, and is placed in a stable position. So, can robots handle mixed cases? Yes, provided the pallet pattern, product data, handling method and upstream flow are designed as one system.
Mixed-case palletising is used where manufacturers or fulfilment operations need to build retailer-ready pallets, customer orders, promotional loads or distribution pallets containing products with different dimensions. It can reduce manual pallet-building work, but it is not a direct replacement for a conventional single-SKU palletising cell. The controls and material handling around the robot are just as important as the robot arm.
Can robots handle mixed cases reliably?
Robotic palletisers can place cases of differing sizes, weights and orientations on the same pallet. A system may build a pre-defined layer pattern, alternate product types within a layer, or construct a fully mixed pallet one case at a time. The suitable approach depends on product range, required pallet stability and order variability.
The simplest application is a limited group of cartons with consistent footprints and known stacking rules. For example, a food manufacturer may combine two carton heights on a standard UK pallet, with heavier cases assigned to lower layers. The robot selects the appropriate handling position and follows a stored recipe for that pallet format.
A more demanding application involves a changing stream of cases, such as an e-commerce or contract packing operation. Here, the control system needs to know the dimensions, weight, permitted orientation and stackability of every case before it reaches the robot. It also needs a confirmed build sequence. If the information is incomplete or cases arrive out of sequence, the robot may be mechanically capable of placing them but unable to build a safe, compliant load.
What makes a mixed-case palletising system work
A mixed-case cell normally combines product identification, case conveying, accumulation, robotic handling and pallet management. The exact layout varies, but the underlying requirement is consistent: the system must match the physical case in front of the robot with the digital instructions for where that case belongs on the pallet.
Case identification may be provided by barcode scanning, label reading, a warehouse management system, manufacturing execution system or a fixed production recipe. Dimensions and weights can be supplied from a product database, verified with checkweighing and dimensioning equipment, or controlled through tightly specified packaging formats.
The pallet-building software then applies rules. These rules determine which cases can be stacked together, where each case can be placed, the maximum permitted height, whether an overhang is allowed, and when layer pads or slip sheets are required. It should also account for the pallet type, as CHEP, Euro and standard UK pallets have different usable footprints and handling requirements.
The most useful design information is usually established before the robot is selected:
- Case length, width, height and weight, including tolerances
- Carton strength, sealing method and permitted top-load
- Product orientation and whether cases may be rotated
- SKU stacking rules, including products that must remain upright
- Required pallet footprints, heights and final load weights
- Order profiles, pallet volumes and changes between shifts
This information is not administrative detail. It determines whether a stable pattern exists and whether the required rate is realistic.
Case quality matters before the palletising point
Robots place cases consistently, but they cannot correct poorly formed or unstable packaging. A case with bowed side panels, inconsistent tape closure or a protruding label can be difficult to grip and can compromise the finished pallet. This is especially relevant when the robot uses vacuum tooling, where a clean, sufficiently flat surface is needed to produce a repeatable grip.
For mixed loads, variation between cartons also matters. A small difference in case height may be acceptable on a single-product pallet but create gaps when combined with other SKUs. Gaps can allow cases to shift during transport. Where variation cannot be removed, pallet-building rules may need to restrict the positions in which that product can be used.
Grippers and end effectors for varied case formats
The end effector is a key decision in mixed-case applications. Vacuum grippers are often suitable for sealed corrugated cases with reliable top surfaces. They can pick a range of box sizes without frequent mechanical adjustment, provided the vacuum zones can be controlled to suit the selected case.
Mechanical clamp grippers can be appropriate where carton surfaces are unsuitable for vacuum or where the case needs a more secure side grip. However, the clamping force must be controlled carefully, particularly for lightweight or partially filled cartons. A clamp that is suitable for one SKU may damage another unless the settings change automatically by recipe.
Some applications use fork-style or hybrid end effectors, particularly for trays, shrink-wrapped bundles or products with inconsistent outer surfaces. The trade-off is often speed, complexity or reduced flexibility. The correct choice should be based on the actual case range and packaging condition, not just the largest and smallest nominal dimensions.
For high-output lines, a multi-pick gripper may lift several cases at once. This is effective where the pallet pattern contains repeated groups of compatible cases. Fully mixed order pallets are more commonly built with single picks because each placement can be different. That can reduce the theoretical cases-per-minute rate, so throughput calculations must include robot travel, scanning, case presentation and pallet changeover rather than pick speed alone.
Pallet stability is the limiting factor
The primary engineering question is not whether the robot can reach every location. It is whether the final pallet will remain stable through stretch wrapping, forklift handling, lorry transport and warehouse storage.
Heavier and denser cases normally belong at the bottom of the load. Cases with low compression strength should not carry heavier products. Similar footprints should be interlocked where possible, while narrow cartons may need to be grouped or restrained by surrounding cases. Some patterns require a layer pad to spread the load or to create a stable surface for the next layer.
A pallet optimiser can calculate space-efficient arrangements, but maximum cube utilisation is not always the best target. A tightly packed arrangement may still be unsuitable if it creates unsupported edges, excessive overhang or a high centre of gravity. Operations teams should agree acceptable limits for pallet height, overhang and weight distribution before the build logic is commissioned.
Stretch wrapping also needs to be considered as part of the palletising process. Film selection, wrap force and top-sheet application must suit the least stable pallet in the product range, not only the most regular one. A well-programmed robot cannot compensate for inadequate load containment after the pallet leaves the cell.
Managing case sequence and buffering
Mixed-case systems depend on sequence. If a pallet recipe requires a heavy carton before a light carton, the heavy carton must be available when needed. In a production line with multiple infeed conveyors, this may require buffering and controlled release of cases. In a fulfilment operation, it may require order sequencing upstream of the palletiser.
There are two common approaches. A fixed sequence system presents cases to the robot in the exact order required for the pallet. This is simpler to control and can suit a stable SKU mix. A dynamic system uses buffers and software to select the next suitable case from several available products. This offers more flexibility but introduces more controls, more exception handling and greater dependence on accurate inventory data.
Exception handling should be planned from the outset. A missing barcode, rejected case, blocked conveyor or unavailable SKU can interrupt a pallet build. The system needs clear rules for whether it holds the pallet, diverts the case, completes a partial pallet or requests operator intervention. These decisions affect both output and traceability.
Where mixed-case robotics is most suitable
Mixed-case robotic palletising is generally most suitable where pallet recipes are repeatable, packaging data is controlled and manual building creates a measurable constraint. It is commonly considered for manufacturer distribution centres, co-packing operations, food and beverage dispatch, pharmaceutical secondary packaging and fulfilment environments handling case-level orders.
It may be less suitable where every order is unique, cartons are highly variable, or the upstream operation cannot provide reliable case identification and sequencing. In those circumstances, a semi-automated station or a process change to standardise outbound pallet formats may deliver a more practical result.
A mixed-case robot should therefore be assessed as a pallet-building system rather than a standalone arm. When case quality, product data, pallet rules and load containment are addressed together, the technology can build varied loads consistently. The most productive next step is to review real order data and a representative sample of cases, then test whether the required pallets are physically stable before setting performance expectations.