Autonomous forklifts and AMRs: evaluating an automation project

The first quote for a pair of autonomous forklifts nearly always arrives with the vehicle datasheet attached: capacity, lift height, speed, battery runtime. Deciding takes three other documents that no supplier can attach. They are the movement history, which sits in the WMS or host system, the dimensioned floor plan with aisles, doors and slopes, and the calendar of busy periods. In many companies that calendar exists only in the shift supervisor's head and has never been written down.

With those three documents on the table you can answer the questions that matter for an investment like this. You find out whether the flows are repetitive enough to support automation, how many vehicles the worst hour needs rather than the average one, which building constraints have to be solved before the order, and which clauses belong in the specification. Many of the problems that surface at start-up come from a figure nobody collected: an underestimated seasonal rush, a gate too narrow to pass at full speed, a floor that becomes a problem at six meters up. The robot is usually the last thing to disappoint.

How many autonomous forklifts do you need? The worst hour decides

Before comparing an AMR with an AGV or an automated forklift, it pays to establish what has to move: which load units, between which points, how often and at what times of day. The mission map shows whether the process is predictable enough to support automation or still so variable that it needs a driver's flexibility. The differences between the technologies are covered in our AGV vs AMR comparison, while the focus here is on method.

A worked example with round numbers shows why the average misleads. A warehouse runs two shifts and moves an average of 30 pallets an hour between dock, storage and lines, but in the first two hours of the morning shift the trucks arrive and missions climb to 55. If the study estimates a full cycle of six minutes for pick-up, travel, drop-off and return, each vehicle completes at most ten missions an hour, before subtracting charging, waits at gates and intersections. On the average, three vehicles are enough. In those two hours it takes six, twice as many, in the same warehouse with the same technology. And a fleet sized on the average would still pass acceptance, if the test were run on a quiet afternoon.

That is why the movement history has to cover a period that truly contains the busiest season, ideally a full year of data recorded by the host system rather than estimated in a meeting. The warehousing literature helps show where to look. The review by de Koster, Le-Duc and Roodbergen in the European Journal of Operational Research (2007), drawing on Tompkins and colleagues, identifies travel as the largest component of order-picking time. The figure concerns picking rather than pallet transport, but the logic carries over. If travel time in your flows is short or too fragmented, no autonomous vehicle will have much to recover, and it should be measured before any quote.

What the feasibility study has to prove

The feasibility study turns these data into a decision that holds up in front of management. The German guideline VDI 2710 on the interdisciplinary design of automated guided vehicle systems, in its June 2025 edition, places it within a path that runs from system identification to detailed design, from purchase to operation, through to modifications and decommissioning. It is a useful map for buyers too, because it is a reminder that decisions made at feasibility stage are paid for over the whole life of the system.

A useful study has to prove at least three things. The first is capacity: how many missions the fleet completes in the busiest hour, with a stated margin well away from the theoretical limit. The second is the number of vehicles, which depends on real cycle times, waits at doors and intersections and how much charging takes away from availability. The third is the financial return, read as total cost over several years. The method for building it, shifts and assumptions included, is in our analysis of AMR fleet ROI, TCO and payback.

There is also something no standard writes for you, the performance acceptance criterion. Safety standards say how the vehicle must protect people. On how many pallets it should move per shift they are silent. That number is up to the buyer, and the study should deliver it already written in testable form, for example missions per hour on a defined flow, measured in an agreed busy window.

When the numbers leave room for doubt, the prudent route is a pilot on a representative flow: a few vehicles, a real circuit, real data to compare with the study. It only makes sense with an exit condition written before starting, valid both ways. If the pilot reaches the planned missions per hour on the chosen flow, it is extended. If it falls short, the layout is redesigned or the project stops, before the rest of the fleet is bought.

Aisles, floor and network: the building constraints the catalog doesn't show

Many projects are decided on physical details. The safety standard for driverless trucks, ISO 3691-4, devotes a normative annex to preparing the zones where the vehicles operate. Along the route it requires, on both sides, a clearance of at least 0.5 m in width and 2.1 m in height from adjacent fixed structures. Where that clearance is missing, or people cannot be detected, the zone becomes an operating hazard zone, with reduced speed and additional audible or visual warnings. For the study this means something precise. Every stretch without clearance is a slow stretch, and a slow stretch on the busiest route eats missions per hour, so the aisle gets measured before the cycle is estimated.

The floor matters most for trucks that lift high, and a little geometry shows why. A 5 mm height difference between the wheels of a vehicle with a one-meter track tilts the mast by five millimeters per meter. At six meters up the forks sit about three centimeters off vertical, just when they have to slide a pallet into the rack slot. Wireless coverage has to be checked along every route, because a network gap is a stop, and doors, freight elevators and fire shutters have to be built into the traffic logic. Charging, finally, is an availability variable and is sized together with the fleet, as explained in our piece on sizing charging stations.

That leaves software integration, often treated as a detail. Vehicles have to receive orders from the host system and report their status, and with vehicles from different makers interoperability becomes a long-term choice. The VDA 5050 specification, published by VDA and VDMA and updated to version 3.0.0 in March 2026, standardizes the exchange of orders and states between a central control and the vehicles. It has no legal force. It covers communication, while the safety conformity of the whole system remains a separate matter, to be assigned by contract to someone who signs for it. Requiring VDA 5050 in the specification costs one line and keeps open the option of growing the fleet with another supplier.

EN ISO 3691-4:2023 and the Machinery Regulation: the standards to write into the specification

The safety of an autonomous handling system is decided at design stage. The technical reference is ISO 3691-4, which sets safety requirements and their verification for driverless industrial trucks and their systems, in its second edition since 2023. It took the place of EN 1525:1997, withdrawn in 2020 when the first edition of the ISO standard came out. Where CE marking is concerned, however, the version that counts is the harmonized European one, EN ISO 3691-4:2023, cited in the Official Journal by Implementing Decision (EU) 2024/1329 of May 13, 2024. That is what gives presumption of conformity with the Machinery Directive 2006/42/EC. Risk assessment under ISO 12100:2010 sits upstream, and safety functions such as person detection, speed control and stopping are designed with the Performance Levels of ISO 13849-1:2023.

If a project assessed today goes into service after January 20, 2027, it will fall under Regulation (EU) 2023/1230, which replaces the Directive from that date. For the first time the Regulation deals with autonomous mobile machinery and defines a supervisory function (Annex III, point 3.2.4). It also requires the control system to perform safety functions on its own even when commands come from remote supervision, and requires movement to take account of the risks of the area where the machine works. The specification should therefore state under which regime the declaration of conformity will be issued and who signs it for the complete system as well as for each vehicle. For groups with plants in the United States the references change: ANSI/ITSDF B56.5-2024 for guided industrial vehicles and the R15.08 series for industrial mobile robots.

The weight of safety in the project, and its return in avoided injuries, is covered separately in our article on AMR safety.

The questions to ask the supplier before the quote

When it comes to comparing suppliers, questions weigh more than datasheets. References are useful if they involve flows similar to yours. A list of installations, on its own, says little. Who takes charge of integration with the host system, and with what responsibilities, should be written down at once, to avoid the back-and-forth between whoever supplies the vehicles and whoever runs the software. The proposed number of vehicles should be justified with your data and, where the route network is complex, with a simulation.

Service and spare parts matter too, as do response times, the option to add vehicles and above all who signs off on the conformity of the whole system when the fleet mixes brands. The topic is developed in our article on automating pallet handling in a running warehouse. A signal to read carefully is a supplier who answers starting from the vehicle model rather than from your flow: it is often a catalog sale with a project label.

Set-up mistakes that get paid for after acceptance

The most frequent stumbles concern method more than technology. The first is automating an inefficient process as it stands: the waste gets written into the software and becomes harder to remove. The second is sizing on the average of the two-shift example, with its twin, an acceptance test run in an ordinary week, which the system passes easily and which says nothing about November. The third is leaving operators out. Without their involvement start-up slows down, and the vehicles get worked around with a few extra manned trucks, which spoils both the statistics and safety. The last is choosing on purchase price, forgetting integration, service and multi-year cost.

The opposite reading holds too. On a single shift, with few missions an hour and routes that change every day, autonomous vehicles spend much of their time idle or charging, and it often pays to reorganize the manual flows first. Saying so at feasibility stage costs little. Finding out after the order costs a fleet.

What to send us for an assessment

A serious supplier asks for numbers on the first call, and nearly all of them are already in the company, scattered across the host system, floor plans and the memory of the people on the floor:

• movement history with time, origin and destination, extracted from the WMS or host system over a period that includes the busiest season;
• hourly profile: missions per shift and the worst hour of the worst week;
• load units: pallets, bins, carts, with dimensions, weights, stability and pick-up and drop-off heights;
• dimensioned layout: aisle widths, slopes, level changes, doors, fire shutters, freight elevators;
• shifts and seasonality, with the campaigns that change the calculation;
• infrastructure: floor flatness, wireless coverage, power points for charging;
• systems to integrate: the WMS, ERP or MES that will assign missions and receive their status.

You can send them by email to info@dipitech.com or on WhatsApp. With these elements the conversation changes level: it stops being about the qualities of the vehicle and starts counting how many missions are needed, in which window of the day and with what margin. The vehicles DIPITECH represents for pallet transport and storage are described on the autonomous robotic forklifts page, part of our industrial automation and material handling robotics offering.