Back to blog
Cold chain5 min read

When should cold storage choose an AMR system, and when AS/RS?

By Dave Hamelink

Short answer

Choose an AMR system when you mainly want to automate internal transport or order picking while keeping your existing building as intact as possible, especially when routes, volumes and SKUs still change regularly. Choose AS/RS when storage capacity, high and stable pallet throughput and a fixed goods flow from inbound to outbound weigh more heavily. It is often not either/or: AS/RS handles the compact storage and AMRs the transport between, and the real question is which processes are stable enough to lock in.

Why has this choice become more relevant?

Automating in a chilled or freezer warehouse puts higher demands on equipment than automating at ambient temperature. Batteries, electronics, charging points, maintenance and the transition between temperature zones all have to be suited to the conditions.

For a long time that counted as a serious objection to mobile robots in freezer environments. By now there are AMR systems developed specifically for temperatures down to -25 °C that can move pallets and roll containers in chilled and frozen environments. That makes an AMR system a serious option for more cold-storage processes. It does not mean every cold store now needs AMRs. It does mean you no longer have to limit the choice automatically to manual transport or a fully fixed automation concept.

The technology is available. The question is whether it fits the problem you are trying to solve.

The AMR-vs-AS/RS decision is not a technology contest; it is a truth test for operational uncertainty.

See how I approach this in the operation.

When does an AMR system fit better?

An AMR system fits best when the bottleneck is in moving goods and not directly in storage capacity. Think of pallets that have to move from receiving to a buffer or pick location, roll containers moving between picking and packing, completed orders going to staging or dispatch, staff who need support during order picking, or transport that has to be automated in an existing warehouse with changing routes.

A key advantage is that this usually needs fewer structural changes to the building. Routes and tasks can be adjusted in software, and capacity can be expanded by adding more robots. That makes an AMR system attractive for organisations that want to automate in phases. You can start with one process or zone, build experience, and expand from there.

But an AMR does not fix a badly designed process.

When replenishment happens too late, master data is wrong, WMS orders are unclear, or load and unload points are constantly blocked, the robot runs into trouble too. It ends up waiting, driving with the wrong priority, or moving goods to a location where no one can process them.

The robot does not make the process better on its own. The process first has to be clear enough to explain to a robot.

When is AS/RS stronger?

AS/RS fits better when the main problem is storage capacity or control over a large, predictable goods flow. That applies, for example, when many pallets have to be stored in a limited floor area, when pallet flows are relatively stable and predictable, when a high baseline throughput is needed, when inbound, buffering, storage and outbound have to function as one whole, and when forklift aisles and manual movements should disappear as much as possible.

In a freezer warehouse, storage density matters even more. The more compactly you can store, the less volume you have to cool. Shuttle AS/RS solutions can also store several pallets deep without separate forklift aisles. With AS/RS, though, the design choices have to be well justified up front: pallet dimensions, product mix, peak volumes, required throughput, maintenance access, redundancy and the desired outbound sequence.

When those assumptions change significantly later, adjustments are usually more drastic than with an AMR system. A different route or task can often be changed in software with an AMR. Extra storage aisles, conveyors, lifts or shuttle capacity cannot be adjusted the same way.

That does not mean AS/RS is always large, rigid or only suitable for new builds. Modern shuttle systems can be built up modularly and integrated into existing or irregularly shaped buildings too. The difference is mainly in how much storage capacity and goods flow you physically lock in up front.

When does a combination make more sense?

In many operations, a combination is stronger than choosing just one technology. AS/RS can handle compact pallet storage, for example, while AMRs handle transport between receiving, the storage system, order picking and dispatch. That way you lock in the stable part of the operation and keep room outside it to adjust routes and capacity.

That is especially attractive when the storage flows are predictable but the processes before or after them carry more variation.

A combination naturally brings extra integration with it. The WMS, WCS or WES has to determine clearly which task is executed by which system, at what priority, and what should happen in a failure. Without proper orchestration, two technically sound systems can still produce a poor goods flow together.

What decision should leadership actually make?

Do not start with a vendor demo. Start with the problem in the operation.

Where is the current bottleneck: storage space, transport, picking, buffering or sequencing? Which volumes and processes have been stable for years, and which change with the seasons, new customers, different SKUs or revised service agreements? How clean are the master data and the orders coming from the WMS? What happens during a peak, a failure or planned maintenance? Can part of the operation keep running when one system is unavailable? And how much capacity must be easy to add as you grow?

With the answers to those questions, the choice becomes a lot more concrete.

An AMR system is not the simple version of AS/RS. And AS/RS is not automatically the more mature solution. An AMR system mainly offers flexibility in transport and picking. AS/RS mainly offers compact storage and control over a fixed goods flow. In many cases, the two complement each other.

So the right choice does not start with the robot. It starts with the process you want to improve and the changes you still expect over the coming years.

Is the operation behaving differently from the plan?

I look at the actual flow: process, system, decision-making, and people. That is usually where the next move is.

More articles