What AMR fleet management is
AMR fleet management is the software, and the computing it runs on, that coordinates a group of autonomous mobile robots (AMRs) working on the same site. Each robot can navigate on its own, but without coordination several robots would compete for the same aisle, queue at the same charger or take jobs that another robot is better placed to do.
The fleet manager sits between the robots and the site’s business systems. It receives work from the warehouse management system (WMS) or manufacturing execution system (MES), breaks it into robot tasks, sends each task to a robot and tracks it to completion. Many people also call it robot fleet management or a fleet management system.
What a fleet manager does
The functions vary by vendor, but most fleet managers cover the same core jobs. The more robots share a site, the more of the value comes from traffic management and task allocation rather than from any single robot.
| Function | What it does | Why it matters |
|---|---|---|
| Task allocation | Matches each job to the best robot by position, battery and load | Shorter travel and fewer idle robots |
| Route planning | Plans paths across the shared site map | Robots avoid known bottlenecks |
| Traffic management | Reserves junctions, aisles and doors, and sets priorities | No deadlocks or head-on standoffs |
| Charging | Sends robots to chargers between jobs | Keeps enough robots available at peak |
| Map management | Holds zones, speed limits and no-go areas | One change applies to the whole fleet |
| Integration | Takes orders from the WMS or MES and reports completion | Robots work to real demand |
| Monitoring | Shows status, faults and performance data | Problems are found before they stop the line |
How the fleet manager talks to robots
Robots and the fleet manager exchange positions, states and orders continuously over the site network, usually industrial Wi-Fi and increasingly private 5G. The fleet manager sends a route or a goal; the robot’s own controller handles local navigation and obstacle avoidance; and the robot reports back its position, battery level and any faults.
Traditionally each robot vendor supplied its own fleet manager, which made mixed fleets hard to run. Open interfaces address that:
-
//
VDA 5050. An interface published by the German automotive and machinery associations VDA and VDMA, using MQTT and JSON messages so one master control can command AGVs and AMRs from different makers.
-
//
MassRobotics interoperability standard. A set of messages that lets robots from different vendors share their position and status, so fleets can avoid each other.
-
//
Vendor APIs. REST or similar interfaces that let a WMS or a third-party orchestration layer send jobs to a single-vendor fleet.
Benefits of coordinated fleets
A well-run fleet manager turns a set of capable robots into a predictable part of the operation.
-
//
Higher throughput. Tasks assigned against real-time congestion.
-
//
Better utilisation. More work from the same robots.
-
//
Safer mixed traffic. Awareness of forklifts and people.
-
//
Resilience. Local control if the network drops.
-
//
Data for planning. Records of travel, waiting and charging time show where layouts or shifts can improve.
Planning a fleet deployment
Most of the work in an AMR project happens around the robots rather than in them. Before scaling up, plan for:
-
//
Network coverage. A wireless survey of every aisle, dock and lift, with roaming tuned so robots do not drop off between access points.
-
//
Site integration. Automatic doors, lifts, fire alarms and conveyors that the fleet manager must signal or wait for.
-
//
Charging strategy. Enough chargers in the right places, and rules for opportunity charging between jobs.
-
//
Zones and rules. Speed limits, one-way aisles and no-go areas agreed with the people who work there.
-
//
A pilot. A small fleet on one process first, then more robots once the traffic rules are proven.
Where fleet management runs
Running fleet management on an edge computer on site keeps decisions fast and independent of the internet. Coordination traffic between dozens of robots is constant and time-sensitive, so most sites keep the fleet manager on premises and use the cloud for reporting and remote support. A compact fanless industrial computer suits smaller fleets, while larger sites use edge AI servers that also process camera feeds.
Adding ceiling-mounted cameras gives the system a view beyond each robot’s own sensors, so it can see forklifts and people approaching a busy corridor and reroute robots before they meet. Our video analytics FAQ explains how those camera streams are processed. On the robots themselves, the robot controller runs navigation and talks to the fleet manager.
The cold-chain case study runs AMR fleet coordination on a Bedrock R8000 edge computer in a refrigerated yard.
Quick answers on AMR fleet management
AMR fleet management: common questions
Is AMR fleet management the same as robot fleet management?
Mostly, yes. Robot fleet management is the wider term and can cover drones or field robots. In warehouses and factories both terms usually mean the same thing.
Can one fleet manager run robots from different vendors?
Increasingly, yes. Interfaces such as VDA 5050 let one master control command robots from several makers, though each vendor’s support varies.
Does fleet management need the cloud?
No. The coordination usually runs on site for speed and resilience. The cloud is often used for dashboards, analytics and remote support.
How many robots need a fleet manager?
Usually two or more. As soon as robots share aisles or chargers, coordination prevents conflicts. Larger fleets depend on it.
What safety standard applies to AMRs?
ISO 3691-4 is the main one. It covers driverless industrial trucks and their systems. The fleet manager supports safety but does not replace each robot’s own safety functions.
We supply edge computers and servers for AMR fleet management and site camera analytics, with the operating system, networking and software stack configured as standard. Tell us the fleet size, site layout and the systems it must connect to, and we will specify the hardware to match.