An AutoStore warehouse can place a large number of small-item inventory bins within a compact grid, but storage density alone does not make it the right fulfillment system. The investment needs to match the operation’s order profile, required throughput, building constraints and labor plan. Start by measuring demand at busy periods, then test whether the proposed grid, robot fleet and workstations can deliver the required flow through picking, packing and dispatch. Compare the full installed and operating cost with suitable alternatives, and include the processes around the system. A design that saves floor space but creates a bottleneck at its ports or replenishment area may not improve fulfillment.
AutoStore is a cube-based storage and retrieval system. Bins are stacked in a grid, and robots move across the grid to access the bins required for an order. The robots bring selected bins to ports, where a worker or another process can pick items. This goods-to-person arrangement can reduce the need for people to walk through storage aisles to collect each item.
The system is one part of a fulfillment operation, not a complete warehouse on its own. It still needs receiving, inventory control, replenishment, picking, packing, exception handling and dispatch processes. The design must also connect with the warehouse management system or other software that directs inventory and orders. A well-designed grid cannot compensate for unclear replenishment rules, poor inventory data or a packing area that cannot keep up.
AutoStore is generally considered for small-item storage, rather than as a substitute for every storage method in a facility. Pallets, bulky products and items that do not suit the selected bin configuration may need separate storage and handling. Confirm product dimensions, weights, handling requirements and inventory rules with the system designer before assuming that a particular SKU can be stored in the grid.
Start with the work the system needs to perform. A useful assessment includes order lines and units per order, the number of active SKUs, demand variation, inventory quantities, replenishment frequency and the proportion of orders that can be picked from the proposed bin assortment. Use actual transaction data covering normal and peak periods, rather than relying on an average day alone.
Look at the relationship between item velocity and order composition. A high-volume SKU that appears in many orders may create repeated demand for its bin, while a broad assortment with low demand per SKU may make inventory location and bin allocation more complex. The system design should account for how stock is replenished and how often bins need to be accessed, not just how many bins fit in the available footprint.
Also identify order types that may follow a different path. Large or fragile goods, mixed orders combining grid-stored and non-grid items, urgent orders and orders requiring special handling can all affect the workflow. Decide whether these orders will be combined at a separate station, routed through another storage area or handled with a different process. The answer changes labor needs and the amount of space required around the system.
Storage capacity and throughput are separate design questions. The grid determines how much inventory can be stored in the system, while the number and configuration of robots, ports, workstations, operators and downstream processes influence how quickly orders can be completed. More storage does not automatically mean more picks per hour.
Ask the designer to model the required flow against the operation’s order profile. The assessment should distinguish sustained demand from short peaks and show the assumptions behind the result. Relevant inputs include how often bins must be retrieved, how many picks are expected at each port, the mix of single-line and multi-line orders, replenishment activity and the time needed for operators to complete tasks. A headline throughput figure is not useful unless its assumptions resemble the warehouse’s actual work.
Ports deserve particular attention. They are points where inventory arrives for picking or other handling, so they can limit the system even if there is spare robot capacity. Consider the number of stations needed, the work each station will perform, the space operators require and how picked goods leave the area. If packing or conveyor handling cannot accept output at the rate picking produces it, the bottleneck simply moves downstream.
Plan for variation, too. Demand may rise during promotions, seasonal periods or operational disruptions. A design intended only for average volume can leave the business short of capacity when it matters. On the other hand, buying capacity for an exceptional peak that occurs rarely may add cost without a sound return. Compare the cost and operational consequences of different peak-handling approaches, including temporary labor, extended operating hours or a separate process for overflow, where those options are practical.
AutoStore should be compared with alternatives that could handle the same inventory and order tasks. The right comparison is not simply which system stores more in a given area. Include picking effort, access, building fit, implementation complexity, future flexibility and the cost of operating the complete workflow.
| Approach | Storage and access | Potential fit | Key limitation to assess |
|---|---|---|---|
| AutoStore | Small-item bins are stacked in a grid and brought to ports by robots. | Operations seeking compact storage and goods-to-person picking for suitable items. | Throughput depends on the full design, including robot activity, ports, staffing and downstream flow. It does not suit every product type. |
| Shelving or pick-face storage | Items are stored on accessible shelves, usually with people moving to the stock. | Smaller operations, varied products or workflows where direct access and simple changes matter. | Travel and aisle space can constrain productivity and storage density as the operation grows. |
| Shuttle-based bin storage | Automated equipment moves bins within a storage structure and presents them to work areas. | Operations evaluating automated bin storage where the system design matches their capacity and flow requirements. | Equipment, layout and operating logic differ by solution; compare the proposed design rather than assuming all systems perform alike. |
| Pallet racking | Pallet loads are stored in rack locations and accessed with suitable material-handling equipment. | Bulk inventory, pallet movements and reserve stock that do not belong in small-item picking bins. | It is not a direct replacement for a high-density small-item picking system; aisle access and picking method affect the comparison. |
These approaches can also work together. For example, a facility may use pallet racking for reserve inventory and an automated bin system for selected small-item picking. This can make sense when each system has a clear role and transfers between them are planned. It can create extra handling if replenishment from reserve stock is frequent or poorly coordinated, so include that movement in the comparison.
Before committing to a layout, check the building conditions that influence the proposed grid and supporting equipment. These may include available floor area and height, columns, floor loading, access for installation and maintenance, utility provision, fire protection requirements and local building or safety requirements. The project team should verify these conditions with the relevant engineers and authorities; assumptions made from a floor plan alone can lead to redesign or additional work.
Map how inventory and orders will move before and after the ports. Receiving and put-away need a workable route to replenishment. Picked items need space for containers, order consolidation or packing, and finished orders need a clear route to dispatch. Include pedestrian and equipment traffic, access for maintenance and safe ways to manage exceptions. A compact storage grid can still sit inside a poorly planned warehouse if adjacent work areas are squeezed or crossing flows create congestion.
Software and data are equally important. Confirm how orders will be released, how inventory will be assigned to bins, how replenishment tasks are triggered, and how the system communicates with existing warehouse and order-management software. Define how staff will deal with inventory discrepancies, damaged goods, short picks, system interruptions and orders that combine stock from different areas. These are operating requirements, not details to leave until after equipment selection.
Request a project-level cost model that covers the full installation and operating life considered by the business. The equipment and grid are only part of the investment. Design and engineering, building work, software and systems integration, installation, commissioning, training and operational changes may also affect the project budget. The exact scope depends on the site and the proposed solution, so ask suppliers to state what is included and what remains the customer’s responsibility.
Operating costs can include labor for picking and replenishment, maintenance, support, energy, software services and spare parts. Also assess the cost of downtime or reduced performance, and the practical plan for keeping orders moving if the automated system is unavailable. Do not assume that automation removes labor: it changes where work happens and which tasks staff perform. The business case should reflect expected staffing by shift and workload, including replenishment, exception handling and supervision.
Compare total costs against a realistic alternative, such as improving the current pick-face layout or using another storage and retrieval method. Use the same order volumes, service requirements and planning horizon for each scenario. Test what happens if demand grows more slowly than forecast, changes in product mix or peaks above the planned level. A payback estimate built on optimistic throughput, immediate labor reductions or unverified savings is not a reliable basis for investment.
This sequence helps expose a mismatch early. If demand data is incomplete, the site is not suitable or the business cannot explain how the system connects to packing and dispatch, resolve those issues before treating a vendor capacity estimate as a final design.
An AutoStore warehouse is worth evaluating when the operation handles a substantial range of suitable small items, has a clear need for compact storage and can benefit from presenting bins to people at workstations. It is more compelling when the anticipated order flow, replenishment process and downstream capacity can all be demonstrated with credible operating data.
Consider another approach, or a combined design, if most inventory moves as pallets, products do not fit the bin strategy, order demand is too low to justify the investment, or the building and workflow make installation difficult. A manual or less automated method may offer a better balance of cost and flexibility for some operations. The decision should follow the task the system needs to perform, not an assumption that higher automation is always more productive.
Before approving a project, ask the supplier and internal team to show how the design meets peak requirements, what assumptions underpin its performance, which costs are excluded and how the warehouse will operate during maintenance or disruption. Those answers are more useful than a density claim in isolation.
The system is designed around bins, so it is most relevant to items that fit the selected bin configuration and can be safely handled and stored there. Product dimensions, weight, packaging and any special storage requirements should be checked during design. Oversized, palletised or otherwise unsuitable stock needs a separate handling plan.
No. It changes the work rather than eliminating every warehouse task. Staff may still receive and replenish stock, pick at ports, pack orders, manage exceptions and support operations and maintenance.
No. Density describes how inventory is stored, not how quickly orders move through the facility. Throughput also depends on the system configuration, port and operator capacity, order mix, replenishment and the ability of packing and dispatch to handle the output.
Not necessarily. Suitability depends on whether products meet the system’s configured storage and handling requirements. Many warehouses need complementary shelving or pallet storage for stock that does not belong in the bin grid.
Request the design assumptions, modeled performance for representative and peak demand, building and integration requirements, project scope, operating-cost breakdown and a plan for maintenance or disruption. Ask how exceptions, replenishment and orders combining different storage areas will be handled.
The soundest AutoStore warehouse decision comes from matching the system to measurable order demand and the complete warehouse workflow. Validate product fit, peak throughput, building requirements and lifecycle costs, then compare the result with credible alternatives. If the design performs under realistic operating assumptions and connects cleanly to replenishment, packing and dispatch, storage density can become a useful part of a wider fulfillment improvement.
