Warehouse automation has improved how goods are stored, retrieved and picked. However, changing assortments, more complex order flows, labour constraints and closer scrutiny of investment are exposing a broader challenge: not simply how to automate more tasks, but how to build an operation that remains manageable as requirements change. Improving individual processes is only part of the answer.
For warehouses, moving from automation towards greater autonomy is not a single technology step. It is a gradual change in how work is divided: systems handle more routine execution and provide better evidence, while teams focus increasingly on judgement, exceptions and improvement. The practical test is whether each step makes the whole operation easier to manage as conditions change.
A useful starting point is the work between automated processes: handovers, product-specific workarounds and adjustments when the order mix changes. These gaps show where systems need broader capabilities, but also where people need clearer information, defined authority or a better escalation route.
Examples from AutoStore’s 2026 Fall Product Announcement illustrate how design, operation and optimisation capabilities can reduce routine coordination in cube-based storage and goods-to-person operations.
Measure the dependency, not just the speed
Throughput matters, but it cannot explain how much coordination an operation requires. Must a supervisor redirect work between stations? Does a different product profile need separate handling? Will a faster process simply move the queue downstream? These questions expose dependencies that headline performance figures can miss.
Shift teams know the workarounds; engineers can assess their effect on the wider flow. Bringing that knowledge into investment discussions gives managers a stronger basis for deciding which interventions to automate and which require human judgement.
Better decisions before go-live
Which assumptions will still hold when demand changes? Design should explore that question before capacity and operating routines become fixed.
AutoStore teams and authorised partners use project benchmarks and AI-assisted comparisons of design scenarios to inform proposals. This evidence can make trade-offs easier to examine, including the risk of installing more capacity than the operation needs. Warehouse teams contribute the local knowledge that a comparison alone cannot establish: service expectations, peak patterns and site-specific constraints.
The useful shift is from accepting a recommended configuration to challenging its assumptions together. Engineers test options against site requirements; operations and finance agree which risks the business can accept. Better analysis should sharpen that discussion, rather than turn an investment decision into an automatic answer.
Make variability easier to manage
Where does everyday variation become an exception? If cases and individual items follow disconnected routes, people may spend their shift reconciling flows instead of managing genuine problems. Automated handling can absorb more of this variation: AutoStore FlexBinsPort handles mixed standard Bin heights at one workstation. AutoCase already combines case and each-picking in one Grid, an announced update will also enable a single task group to operate across multiple stations in parallel. The relevant measure is whether these capabilities remove routine coordination from the wider flow.
The operational question is what happens next. Packing capacity, replenishment and escalation routes need to fit the wider flow. As systems absorb more variation, supervisors should have more scope to resolve unusual conditions and improve routines, instead of repeatedly compensating for gaps. Training should reflect that change, including when to intervene and what evidence to record.
Turn visibility into informed action
A report can show where attention is needed without explaining which response makes operational sense. Teams need to connect system evidence with the order profile, recent changes and service priorities. Otherwise, more data risks becoming another task rather than a basis for improvement.
CubeFleet, announced as part of AutoStore’s Fall portfolio, will provide usage-based maintenance visibility through thresholds, alerts and guidance. It shows why visibility alone is not enough: earlier awareness of service needs can support maintenance planning, but the team still decides when an intervention is justified and how it should be prioritised alongside fulfilment requirements. Recording what changed and checking the result closes the loop, allowing experience to inform the next decision.
A clearer brief for systems and teams
The next investment brief should define the division of work as clearly as the equipment: map repeated interventions, specify which the system will handle, and agree how teams will recognise and escalate exceptions. Include maintenance and training from the outset.
“The useful question is which routine tasks the system can take on, and where the team needs better information to make a sound judgement. Progress should make those responsibilities clearer.” Russell Holmes, UK&I sales director, AutoStore.
A practical measure of progress is not automation alone, but a reduction in avoidable interventions combined with a team that is better equipped to improve the operation. Greater autonomy develops when systems can execute a broader range of routine work, while people have the information and authority to manage exceptions and adapt the operation as requirements change.
Comments are closed.