You’ve decided autonomous forklifts are the right move for your facility. The business case is clear: lower labor costs, 24/7 operation, fewer accidents, and predictable throughput. But then comes the question that stops most projects in their tracks: how many autonomous forklifts do you actually need?

Order too few and your automation investment becomes a bottleneck, frustrating workers and failing to deliver the ROI you promised leadership. Order too many and you’ve tied up capital in robots that spend half their shift parked and waiting. Getting autonomous forklift fleet sizing right is one of the most consequential decisions in any warehouse automation project—and it’s one that too many teams approach with guesswork rather than methodology.

This guide gives you a practical, structured framework for calculating the right fleet size for your specific operation, covering everything from throughput requirements and cycle times to shift patterns, charging windows, and the type of autonomous forklift your tasks demand. Whether you’re automating a single zone or planning a full digital factory transformation, the principles here will help you arrive at a number you can defend—and a deployment you can scale.

Why Getting Fleet Size Right Matters More Than You Think

Fleet sizing isn’t just a procurement question—it’s the foundation of your entire automation architecture. An undersized fleet creates cascading delays: when one robot is charging, another is handling an exception, and a third is navigating around a blocked aisle, your throughput drops below what you achieved with manual forklifts. That outcome erodes trust in the technology and makes it nearly impossible to build internal support for further automation investment.

An oversized fleet brings its own problems. Autonomous forklifts require infrastructure: charging stations, defined travel lanes, integration with your Warehouse Management System (WMS), and floor space that isn’t occupied by other equipment or inventory. Too many robots in a confined space actually reduces efficiency because the fleet management system spends more time resolving traffic conflicts than moving loads. There’s also a straightforward financial argument: autonomous forklifts represent a significant capital investment, and machines that sit idle represent money that could have been deployed elsewhere.

The good news is that fleet sizing follows a logical methodology once you understand the variables involved. It’s an engineering problem, not a guessing game, and the companies that approach it systematically consistently achieve faster payback periods and smoother scale-up trajectories.

The Key Variables That Drive Fleet Size

Before you can calculate a number, you need to gather accurate data across several operational dimensions. Skipping or estimating any of these inputs introduces error that compounds quickly. The variables you must quantify are:

• Daily pallet moves (or equivalent unit loads): Total number of transport tasks required per operating day, broken down by shift if your volume is uneven.
• Average cycle time per task: The time from when a robot picks up a load to when it returns to a ready state, including travel to pickup, load engagement, travel to drop-off, unloading, and return travel.
• Operational hours per shift: The number of hours the fleet is expected to be actively moving loads, not including breaks, charging windows, or maintenance periods.
• Number of shifts per day: Single-shift, double-shift, and 24/7 operations have dramatically different fleet size implications.
• Charging time and battery autonomy: Most autonomous forklifts in the 1,500–3,000 kg class deliver 6 to 10 hours of runtime per charge. Opportunity charging during breaks can extend effective uptime, but requires planning.
• Task variability and peak demand: Average throughput is rarely what sizes a fleet. Peak demand windows—end-of-shift surges, inbound truck arrivals, promotional periods—determine your practical ceiling.
• System availability and maintenance buffer: No fleet operates at 100% uptime. A realistic availability assumption of 90–95% means you need to build buffer capacity into your calculation.

Gathering this data from your current operation—even if it’s manually driven today—gives you the baseline from which all fleet sizing math flows.

A Step-by-Step Framework for Calculating Your Fleet Size

With your operational data in hand, the calculation itself follows a straightforward sequence. Here’s how to work through it systematically.

Step 1: Calculate Required Cycles Per Hour

Divide your total daily task volume by the number of productive operating hours available. For example, if you need to move 480 pallets across an 8-hour shift, you require 60 pallet moves per hour. If your peak window compresses 40% of that volume into 3 hours, your peak rate is actually closer to 64 moves per hour—and that peak figure, not the average, should drive your sizing.

Step 2: Determine Cycles Per Robot Per Hour

Measure or model the average cycle time for a single robot on your most common task routes. A robot with a 12-minute average cycle time completes 5 cycles per hour. Longer routes, multi-stop missions, or facilities with significant elevation changes (ramps, mezzanines) will extend cycle times and reduce per-robot output. This is where accurate facility mapping and route simulation pay off.

Step 3: Calculate the Raw Fleet Size

Divide the required cycles per hour by the cycles per robot per hour. Using the example above: 60 required cycles divided by 5 cycles per robot equals 12 robots as a theoretical minimum. This is your baseline number before any adjustments.

Step 4: Apply the Availability and Charging Buffer

Multiply your raw fleet size by a buffer factor to account for charging rotations and planned maintenance. A commonly used buffer of 1.15 to 1.25 (representing 80–87% effective fleet availability at any given moment) brings our example to 14–15 robots. If your operation runs 24/7 with no charging windows during breaks, you’ll need to model charging shifts explicitly, which often pushes the buffer factor higher.

Step 5: Validate Against Your Facility Constraints

Run a final sanity check against your physical space. Can your aisle width and traffic lane design support that number of robots operating simultaneously without excessive conflict? Is there sufficient charging infrastructure? Does your WMS or fleet management platform have the license capacity and processing headroom to orchestrate the fleet at that scale? If any of these constraints are binding, you may need to either expand infrastructure or revisit your throughput targets.

Hidden Factors Most Planners Overlook

The five-step framework above gives you a solid number, but experienced automation planners know that several additional factors can shift that number significantly in either direction. Ignoring them is one of the most common reasons early-stage autonomous forklift deployments underperform.

Traffic interaction with manual equipment: If your facility operates a mixed fleet—autonomous forklifts alongside human-driven equipment or pedestrian traffic—the AMRs will slow down or reroute more frequently than they would in a fully autonomous zone. This effectively extends cycle times and reduces per-robot throughput. Factor in a 10–20% cycle time penalty for high-traffic mixed environments.

Load type variability: Autonomous forklifts are optimized for consistent, standardized loads. If your operation handles irregular pallet configurations, oversized loads, or goods that require manual verification before pickup, exception handling time adds to each cycle. Quantify your exception rate and build it into your cycle time estimate.

Seasonal and promotional demand spikes: A fleet sized for average annual throughput will be overwhelmed during peak periods. If your peak demand is 40% above average (common in retail supply chains and e-commerce fulfillment), size for the peak and accept that utilization will be lower during off-peak periods. The alternative—renting manual forklifts during peaks while running an autonomous fleet during base periods—defeats much of the purpose of automation.

Growth trajectory: Automation infrastructure has a long lifespan. A fleet sized for today’s throughput may be inadequate within 18 months if your business is growing. Build a 2–3 year volume projection into your sizing discussion and confirm that your chosen platform can scale incrementally without requiring a full system redesign. Reeman’s autonomous forklift lineup, for example, is designed for modular deployment—you can add units to an existing fleet as volume grows, with the same SLAM navigation and fleet management platform managing the expanded operation seamlessly.

Does the Type of Autonomous Forklift Change the Math?

Absolutely—and this is a dimension that often gets collapsed into a single ‘autonomous forklift’ category when it deserves much more nuance. Different forklift types have different throughput profiles, payload capacities, and use-case fits, which means the same task profile might require a different number of units depending on which model you deploy.

For heavy-duty pallet transport in wide-aisle environments, a high-capacity counterbalance autonomous forklift like the Reeman Rhinoceros Autonomous Forklift handles substantial payloads with high travel speeds, meaning fewer units may be required to hit your throughput target. For racking operations requiring precise fork positioning at height, a reach-type autonomous forklift such as the Reeman Ironhide Autonomous Forklift brings specialized capability—but its cycle time profile in a high-bay environment will differ from a floor-level transport task.

For tighter spaces or lighter-load applications, compact autonomous stackers like the Reeman Stackman 1200 offer excellent maneuverability in narrow aisles where larger units would create traffic conflicts. In many facilities, the optimal answer isn’t a homogeneous fleet of one model—it’s a mixed fleet where each robot type handles the tasks it was designed for, with the fleet management system orchestrating task assignment dynamically.

This task-to-robot matching exercise can meaningfully reduce your total fleet count compared to sizing with a single generalist unit, while also improving overall system efficiency.

Start With a Pilot, Then Scale With Confidence

Even the most rigorous fleet sizing calculation is, at its core, a model built on assumptions. Real facilities have quirks—floor conditions, workflow patterns, employee behaviors, and integration edge cases—that no model fully captures. This is why the most successful autonomous forklift deployments almost universally begin with a structured pilot phase rather than a full fleet rollout on day one.

A pilot deployment of 2 to 5 units in a well-defined operational zone gives you ground-truth data on actual cycle times, exception rates, charging patterns, and system integration performance. That data then feeds back into your fleet sizing model, allowing you to refine your projections before committing to full-scale procurement. It also builds internal familiarity with the technology among your warehouse team, which accelerates adoption when the broader rollout begins.

Reeman’s plug-and-play deployment philosophy is specifically designed to support this phased approach. With laser navigation and SLAM mapping that doesn’t require facility modifications or reflector infrastructure, robots can be operational in a new zone quickly. The same fleet management platform scales from a 3-unit pilot to a 30-unit full deployment without requiring a system re-architecture, protecting your initial integration investment as your fleet grows.

After your pilot, revisit the five-step calculation with real data replacing your initial estimates. You’ll typically find that cycle times are slightly different from projections, charging patterns need adjustment, and one or two task types benefit from a different robot model than originally planned. This iteration cycle is normal and healthy—it’s how you arrive at a fleet configuration that performs in the real world, not just on paper.

Conclusion

Autonomous forklift fleet sizing is part science, part operational judgment. The science gives you a defensible baseline: calculate your required throughput, model your cycle times, apply realistic availability buffers, and validate against your facility constraints. The operational judgment comes in when you account for peak demand variability, mixed-fleet dynamics, load type complexity, and your business growth trajectory.

The most important takeaway is that fleet sizing is not a one-time decision made at project kickoff—it’s an ongoing process. Start with a structured pilot, measure real-world performance, refine your model, and scale incrementally. This approach consistently delivers better outcomes than committing to a large fleet based on theoretical calculations alone.

If you’re planning an autonomous forklift deployment and want to apply this framework to your specific operation, Reeman’s engineering team works with facilities across manufacturing, logistics, and warehousing globally to develop fleet sizing recommendations grounded in real operational data. With models ranging from compact stackers to heavy-duty counterbalance forklifts, the right configuration for your throughput needs is likely closer than you think.