AMR planning guide
How to Calculate AMR Fleet Size: Throughput, Cycle Time, and Charging
Build a defensible robot-count estimate from the work, then test the assumptions that can change the answer.
The right AMR fleet size is the smallest fleet that meets the required task demand under the real route, handoff, charging and recovery conditions. Start with a time-phased workload and a complete cycle time; do not divide a daily move count by a brochure speed.
Start with demand
Measure the work before you count robots
Create a task list for a representative busy period, including shift changes and exception work. For each task, record the load, start and end point, due time, handoff method and whether a person or another system must confirm completion.
| Input | What to capture |
|---|---|
| Demand | Tasks by 15- or 30-minute interval, shifts, peaks, due times and priority jobs. |
| Cycle | Travel, pickup, drop-off, waiting, handoff confirmation and recovery time for a complete task. |
| Route | Distance, crossings, doors, lifts, congestion, floor conditions and one-way rules. |
| Availability | Charging, planned maintenance, network interruptions, blocked routes and manual recovery. |
| Service target | Moves per hour, queue limit, late-task tolerance and the reserve required for the operation. |
Calculation
Use complete cycle time and productive availability
For a first estimate, calculate the robot-hours required in each planning interval, then divide by the productive hours available from one robot in that interval. A practical worksheet can be written as:
Fleet estimate = (tasks × complete cycle time) ÷ productive time per robot, then adjust for peak demand, charging, availability and the agreed reserve.
Complete cycle time includes travel in both directions when relevant, pickup and drop-off, queueing, handoff confirmation and the recovery time you observe in the process. Productive time excludes charging, planned maintenance, blocked routes, software recovery and other non-productive periods. Use the same units throughout and calculate each interval separately before summing the result.
Do not hide the reserve inside a mysterious multiplier. State what it protects: a demand spike, one robot unavailable, a long charger queue, a service-level target or future growth. This keeps the model explainable when operations challenges the result.
System constraints
Look for the bottleneck that a larger fleet cannot fix
More robots do not solve a single narrow door, slow workstation, shared lift, limited charger or congested crossing. Map queues and handoffs alongside robot travel. If all robots wait for one resource, the resource—not the fleet count—is setting throughput.
- 01Compare task demand with charger capacity and the battery window during the busiest shift.
- 02Check whether doors, lifts, conveyors, racks or workstations create a queue longer than robot travel.
- 03Model blocked paths, manual recovery and priority jobs instead of assuming a perfect map.
- 04Separate an increase in fleet availability from an increase in physical throughput.
Pilot and acceptance
Validate the count against a measured operating scenario
Use a representative route and workload to compare the estimate with observed moves, queue time, charging behavior, exception recovery and on-time completion. Test the busiest sustained interval, a normal interval and at least one degraded condition such as a blocked route or unavailable robot.
Record the assumptions that remain uncertain: future task growth, interface latency, pedestrian density, battery aging, maintenance windows and the reserve policy. Tie the final fleet recommendation to a measurable service target and review it when the workflow changes.
For a broader project brief, use the AMR cost guide to separate robot, integration, site and support scope, then compare the result with the relevant AMR platform and application requirements.
Buyer questions
FAQs
What is the basic AMR fleet-size formula?
A useful starting point is required robot-hours per planning period divided by productive hours available per robot, then adjusted for charging, availability, peak demand and a documented reserve. Validate the result with a pilot or simulation.
Should I size from average moves per hour?
No. Average demand can hide peaks, shift changes and blocked routes. Use a time-phased demand profile and identify the busiest sustained interval the fleet must serve.
How do charging and battery swaps affect fleet size?
Charging removes productive time and may also create congestion. Measure charge duration, usable battery window, charger capacity and the effect of opportunity charging or spare batteries on availability.
How much spare capacity should an AMR fleet have?
There is no universal percentage. Define the service target, failure response, maintenance plan and peak profile, then choose and document a reserve that your acceptance test can verify.
Continue your evaluation
Need help sizing the first workflow?
Share task volumes, payloads, routes, handoffs and charging conditions to turn a robot-count guess into a testable project scope.
Talk to an AMR specialist