Autonomous mobile robots in the warehouse: when it makes sense and when it does not
How to assess warehouse readiness for AMR deployment - without the marketing simplifications that robots will fix everything.
Autonomous mobile robots for the warehouse - AMRs - are no longer exotic. Several vendors offer solutions across different price ranges, implementation cases are accumulating, and the barrier to entry is coming down.
At the same time I see two common patterns of mistaken decisions. The first: buying robots too early, before the process is ready. The second: deferring indefinitely, waiting for perfect conditions. Between them is the zone where the decision makes sense.
What AMRs do well
AMRs solve a specific problem: moving goods or containers within a warehouse without continuous human involvement. They navigate independently, avoid obstacles, return to charging, and integrate with the warehouse management system.
The core economics of AMRs is reducing personnel travel distance. In large warehouses, operators walk tens of kilometres per shift. If a robot handles transportation, the person focuses on tasks that require hands: picking, packing, quality control. Productivity grows not by speeding up each individual operation, but by redistributing labour.
When AMRs will not solve the problem
Robots do not compensate for a chaotic process. If a warehouse operates poorly because goods are placed unpredictably, bin locations are not maintained, and tasks are created manually with errors - robots will add a new complex element to that chaos.
AMRs require stable infrastructure. They need a predictable floor plan - without constant rearrangement of racks, with adequate aisle widths, with working Wi-Fi across the entire navigation area. In a warehouse with an unstable layout or poor wireless coverage, this will become a source of constant incidents.
WMS integration is critical. Without a proper warehouse management system, robots do not know where to go or what to carry. If the company has no WMS or it operates unreliably, that needs to be fixed first.
What to look at when evaluating
Before deciding on a pilot, I look at a few parameters.
Volume of internal movements. How many movement tasks per shift? If it is a handful, robots will not generate savings. If it is hundreds or more, the economics are worth calculating.
Layout stability. Has the warehouse layout changed in the past six months? Will it change in the next one or two years? An unstable environment means constant reconfiguration of robots.
WMS maturity. Does the system have current bin-level stock, movement tasks, and operation history? Without this, automation is simply not possible.
Personnel readiness. AMRs change the role of operators but do not eliminate them. You need people who are ready to work alongside robots and handle their technical needs.
Questions to assess readiness
- How many movement operations per shift, and what share of them can be standardised?
- Is there a WMS with current data at bin level?
- Is the layout stable - will it not change radically over two years?
- Is there Wi-Fi coverage across the entire working area?
- Who will own the robot fleet - who makes decisions on maintenance and integration?
If most questions have clear answers, a pilot makes sense. If not, investing in robots without first addressing the process will be disappointing.