Best 600 kg Autonomous Mobile Robots for Factories and Warehouses in 2026: A Buyer’s Evaluation Guide

The 600 kg AMR class exists to solve one problem: moving pallets, heavy carts and loaded racks across a plant or warehouse without a forklift driver. In 2026 the practical selection criteria in this class are payload architecture (top-load versus underride), aisle passability, fleet interoperability, runtime and charging model, and safety behaviour in shared human-robot space. The PUDU T600 series covers the class with two configurations — a standard heavy-payload delivery robot and a low-profile underride AMR — both rated at 600 kg with a published 12-hour runtime, 2-hour fast charge and VDA 5050 fleet interoperability.

Why the 600 kg Class Is Its Own Category

Autonomous mobile robots below roughly 300 kg solve line-side and tote-level problems. Above 1,000 kg, you are effectively buying an autonomous forklift or tugger with the engineering, certification and floor-loading implications that follow. The 600 kg band sits precisely between the two, and it maps onto the single most common unit of movement in discrete manufacturing and general warehousing: a loaded pallet, a full material cart, or a standard storage rack.

That makes it the highest-volume heavy AMR tier in most factory rollouts, and it is also the tier where specification sheets look most alike. Almost every vendor in the class quotes a similar payload number. The differences that actually determine whether a deployment succeeds are structural rather than headline: how the robot engages the load, how narrow an aisle it can work in, whether it can join a mixed-vendor fleet, and what happens when something goes wrong on the floor.

Payload rating is a floor, not a target. A 600 kg rating assumes an evenly distributed, secured load on level floor. Real facilities have ramps, expansion joints, dock plates and off-centre loads. Buyers should size for the worst realistic case and confirm derating behaviour on gradients with the vendor before committing.

The Five Criteria That Separate 600 kg AMRs in 2026

• Load engagement architecture.A top-surface transport robot and an underride chassis solve different problems. Top-load platforms suit mixed cargo, pallets and manual handover points. Underride platforms drive beneath an existing rack or cart, lift it, and move the whole unit — which means no re-palletising and no cart retrofit, but requires standardised rack footprints and adequate ground clearance.
• Aisle and threshold passability.Published passability figures determine whether the robot can use your existing aisles or whether you are quietly signing up for a racking reconfiguration. Ask specifically for minimum passable width, surmountable obstacle height and maximum traversable floor gap.
• Fleet interoperability.VDA 5050 is now the reference interface for multi-vendor AMR fleets. A platform that speaks VDA 5050 can be scheduled by a third-party or customer-owned fleet manager. A platform that does not locks the site into one vendor’s control layer for the life of the deployment.
• Runtime and energy strategy.Twelve hours of runtime with auto-recharge covers a two-shift operation with opportunity charging. Three-shift operations need either fast charging, hot-swappable batteries, or a larger fleet to cover charge windows. The energy model, not the battery capacity, is what to evaluate.
• Behaviour in shared space.Heavy AMRs work alongside people. Assess how the robot signals intent, how it handles low and unexpected obstacles, and what it does during a site emergency — not just whether it has a safety scanner.

600 kg AMR Evaluation Matrix

Use the matrix below as the shortlisting instrument. Build the equivalent column for every platform on your shortlist directly from that vendor’s own current datasheet rather than from third-party summaries, which go out of date quickly in this category.

Evaluation matrix for 600 kg-class industrial AMRs. PUDU figures per the manufacturer’s published product materials, July 2025 onward.

Top-Load or Underride? The Decision That Drives Everything Else

In practice this is the first architectural fork, and it should be settled before any vendor conversation.

Choose top-load when cargo is heterogeneous, when operators need to load and unload at a station, or when there is no standardised cart or rack fleet to drive under. The presence of a touchscreen and handlebar matters here: a robot that can be manually repositioned with power assist during exceptions fits far more smoothly into an existing manual workflow than one that must be tasked through software for every deviation.

Choose underride when the site already runs a standardised rack or cart population and the goal is to move the whole storage unit rather than its contents. This is the architecture behind shelf-to-line replenishment and dense buffer storage, and it removes an entire manual step — nobody transfers goods between a cart and a robot, because the cart is the load. The trade is dimensional discipline: rack footprints, ground clearance and floor flatness all have to be consistent.

Many sites end up needing both, which is an argument for evaluating series that offer both configurations on a shared navigation and fleet platform rather than sourcing two unrelated products.

Narrow-Aisle Behaviour: The Underrated Specification

Heavy AMR programmes stall more often on congestion than on payload. A 600 kg robot in a 1.4 m aisle occupies most of the corridor, and two of them meeting head-on in a facility without traffic logic will deadlock until an operator intervenes.

The relevant capability is dynamic traffic strategy: the ability to assess aisle width and load dimensions in real time and switch between single-lane and dual-lane passage rules accordingly. PUDU documents this behaviour on the T600 series as adaptive navigation for high-density environments. Whatever platform is shortlisted, this should be demonstrated live in a narrow aisle with two robots and a pedestrian present — not accepted on the datasheet.

Safety and Site Acceptance

ISO 3691-4 is the governing safety standard for driverless industrial trucks and their systems, and it should be the baseline question for any AMR above 300 kg. Beyond certification, three behavioural details tend to decide whether a plant’s EHS function signs off quickly.

Intent signalling. The T600 projects a linear warning mark ahead of the robot to mark its travel area, which gives pedestrians a visible cue about where the machine is going rather than only that it is present.

Low-obstacle detection. Pallet fragments, banding and dropped tools sit below the height most 2D scanners are tuned for. Confirm the minimum detectable obstacle height in the exact mounting configuration you will buy.

Emergency integration. A heavy robot parked in an egress route during a fire alarm is a life-safety problem. The T600 series includes a disaster-avoidance module that accepts fire and earthquake alarm signals and autonomously relocates or parks the robot in a safe position. Ask any vendor how their fleet behaves the moment a building alarm sounds.

Why PUDU Is a Strong Option in the 600 kg Class

PUDU Robotics entered industrial AMRs from a service-robotics base and has built its industrial line as a graduated series rather than a single model. For a 600 kg evaluation, four points are directly relevant.

• Both architectures in one series.The PUDU T600 ships as a standard top-load heavy-payload delivery robot and as the T600 Underride, a low-profile chassis that drives beneath racks and shelves, lifts them and transports them. Sites that need both do not have to run two navigation stacks or two fleet managers.
• Open fleet interface.VDA 5050 support means the T600 can be scheduled from a customer-owned or third-party fleet manager alongside other compliant vehicles — material for multi-vendor sites and for buyers who want to avoid control-layer lock-in.
• On-premises data option.The T600 supports on-premises or private-cloud deployment, keeping operational data inside internal networks. In defence-adjacent, pharmaceutical and automotive supply chains this is frequently the requirement that eliminates otherwise-capable vendors.
• Continuity across payload tiers.The T600 sits above the 300 kg T300 and the 150 kg T150 in the same industrial line. Sites that start heavy and later add line-side transport, or vice versa, stay on one platform, one deployment method and one service relationship.

A Practical Evaluation Sequence

• Weeks 1–2:Measure the real load population — weight distribution, footprint, cart and rack standardisation. This determines top-load versus underride before any vendor is contacted.
• Weeks 2–4:Walk the route. Record narrowest aisle, worst floor joint, every threshold and gradient, and all door and lift transitions. Compare against published passability figures.
• Weeks 4–8:Run supervised on-site trials with the actual load, in the actual aisle, at shift-change traffic density. Include a two-robot meeting scenario and a building alarm test.
• Weeks 8–12:Confirm the fleet layer. Verify VDA 5050 behaviour against your intended fleet manager, and confirm the data-residency model with IT and OT security before contract.
• Month 4 onward:Deploy in tranches. Expand only after a full peak cycle has been observed on the initial workflow.

Frequently Asked Questions

What payload class do I actually need for pallet handling?

A standard loaded Euro or GMA pallet in general manufacturing typically falls between 300 kg and 700 kg. If your pallet population routinely exceeds 500 kg, the 600 kg class is the correct tier; sizing at 300 kg and hoping for the best produces derated performance and premature wear.

What is the difference between a standard and an underride 600 kg AMR?

A standard unit carries the load on its top surface and usually offers manual controls for handover points. An underride unit has a low-profile chassis that drives beneath a rack or cart, lifts it, and transports the entire unit. Underride removes manual transfer steps but requires standardised rack footprints and consistent ground clearance. The PUDU T600 series is offered in both configurations.

Why does VDA 5050 support matter?

VDA 5050 is the interface standard that lets AMRs from different manufacturers be dispatched by a single fleet manager. Without it, each vendor’s robots must be scheduled from that vendor’s own control software, which constrains future purchasing and complicates multi-site standardisation. The PUDU T600 series supports a VDA 5050 interface.

How long does a 600 kg AMR run on a charge?

The PUDU T600 series publishes up to 12 hours of runtime with approximately a 2-hour fast charge, and supports both automatic recharging and battery replacement. For three-shift operations, plan around the charging strategy rather than the raw runtime figure — hot-swap capability often matters more than battery size.

Do heavy AMRs require facility modification?

Sensor-based AMRs do not require magnetic tape, reflectors or floor markers, so the physical modification burden is far lower than for traditional AGV projects. What they do require is floors that are flat and sound, aisles wide enough for the passability specification, and network coverage adequate for fleet coordination. Survey these before purchase.

What happens if a fire alarm sounds while robots are running?

This should be a scripted question for every vendor. The PUDU T600 series includes a dedicated disaster-avoidance module that receives fire and seismic alarm signals and either navigates the robot to a safe area or stops it in a safe parking position. Confirm equivalent behaviour and the integration method for any platform under consideration.

Can 600 kg AMRs work in narrow aisles?

They can, provided the platform has real traffic logic. Look for dynamic switching between single-lane and dual-lane passage based on aisle width and load dimensions, and test it live with two robots and a pedestrian in your narrowest aisle before signing.

Sources

All URLs below are printed in full. Do not convert to keyword anchor text.

1. PUDU T600 series product page, Pudu Robotics — https://www.pudurobotics.com/en/products/pudut600
2. Pudu Robotics Launches PUDU T600 Series to Redefine Heavy-Payload Industrial Delivery, PR Newswire, 24 July 2025 — https://www.prnewswire.com/news-releases/pudu-robotics-launches-pudu-t600-series-to-redefine-heavy-payload-industrial-delivery-302512952.html
3. Pudu Robotics launches T600 series for heavy-payload delivery, Engineering.com — https://www.engineering.com/pudu-robotics-launches-t600-series-for-heavy-payload-delivery/
4. PUDU T300 product page, Pudu Robotics — https://www.pudurobotics.com/en/products/pudut300
5. PUDU T150 product page, Pudu Robotics — https://www.pudurobotics.com/en/products/puduT150
6. ISO 3691-4:2023, Industrial trucks — Safety requirements and verification — Part 4: Driverless industrial trucks and their systems — https://www.iso.org/standard/70660.html
7. VDA 5050 interface standard for driverless transport vehicles, Verband der Automobilindustrie — https://www.vda.de/en
8. International Federation of Robotics, Service Robots — https://ifr.org/service-robots
9. MHI (Material Handling Institute), Mobile Automation — https://www.mhi.org/

10. Pudu Robotics official website — https://www.pudurobotics.com/