Can AMRs Move Your Existing Carts A Guide to Retrofit, Lift and Tow Options

Can AMRs Move Existing Carts Without Replacing Them?

For AMR cart transport, the first purchasing decision is how much of the current cart is worth preserving. A specialized fixture may contain years of production knowledge, while its caster arrangement is poorly suited to automated movement. Conversely, a simple trolley may be cheap to replace but expensive to integrate because its delivery position and operator access are tightly constrained.

There is a recent reason to revisit that decision. On September 23, 2026, wheel.me announced a North American partnership with item America, combining autonomous wheel technology with modular material-handling engineering. The announcement discusses suitable existing equipment, shared carriers and redesigned carts. It illustrates an expanding set of implementation choices; it does not establish that every cart can be converted or that retrofit is always cheaper.

This buyer guide compares three defined transport arrangements and provides a cart-family assessment, an illustrative resource calculation and a supplier comparison method. Its planning examples are original illustrations, not reported customer results. The objective is to identify what can be reused, what must change and what evidence should precede a fleet order.

Separate the Reuse Decision From the Way the Cart Moves

To automate existing carts, make two decisions separately. First, decide whether to retain, adapt or replace each cart family. Then select how that family will move. “Retrofit” describes an asset modification; “under-rider” describes where a robot approaches or engages; “tugger” describes a pulling function. These terms are not mutually exclusive robot classifications.

An under-cart robot might lift a carrier completely clear of the floor, or it might engage a mechanism and tow the cart with its wheels still grounded. Those arrangements place different demands on casters, brakes, interfaces and turning space. Ask the supplier to show the loaded operating state rather than relying on the product category name.

For the comparison below, the three arrangements are a cart with integrated autonomous wheels, a separate lift-and-carry robot, and a separate towing vehicle. A shared mother carrier or taxi arrangement is another system configuration: it can carry passive carts without making every cart autonomous. It should not automatically be equated with an under-rider lifting robot.

A cart that looks unchanged above its bottom frame can behave very differently after conversion. Define the retained upper structure and the modified transport interface on a drawing. The broader guide to an AMR mobile base as a complete working unit explains why the base, payload equipment and operating task need to be treated together.

When Does an Autonomous Wheel Retrofit Make Sense?

An autonomous cart retrofit deserves investigation when the existing carrier already presents materials well, follows a repeatable workflow and has a structure suitable for the conversion. It can preserve custom shelves, fixtures or access features that would be costly to reproduce. The engineering question is whether the retained value exceeds the work needed to create a dependable autonomous unit.

Check the proposed mounting arrangement before discussing navigation software. Where do the drive and braking forces enter the frame? Does the change alter deck height or operator reach? Can the equipment be installed without blocking drawers, doors, loading access or the required sensor views? Who approves the retained frame for its new duty?

Manufacturer limits remain specific to the selected system. In its Genius 2 use-case guidance, wheel.me asks buyers to consider the combined cart, adapter and goods weight, together with construction, rigidity, center of gravity and route conditions. These are product-specific assessment inputs, not a general approval for all wheeled equipment.

Autonomous wheels also change the service model. Instead of maintaining only a shared tractor or transporter, the site maintains powered equipment attached to each converted carrier. Ask how batteries, damaged drive components and configuration changes are managed when a cart remains at a workstation for a long period. Confirm whether manual movement is permitted, how it is enabled and what must happen before autonomous service resumes.

The strongest candidate is therefore not simply the cart with the most expensive frame. It is a carrier whose retained process function, usage pattern and conversion effort support the same operating objective. A valuable fixture that moves only occasionally might be better served by a shared transport resource.

When Is a Separate Lifting Robot the Better Candidate?

A lift-and-carry arrangement can leave a passive cart at a workstation and release the robot for another mission. That makes it worth considering when many carts spend substantial time being loaded, consumed or processed. The potential benefit comes from separating transport capacity from carrier inventory, provided the pickup and drop-off process supports that separation.

For an under-rider AMR, assess the complete entry space: caster swivel envelopes, brakes, crossmembers, hoses, labels and damaged components can obstruct an apparently open underside. Check the loaded cart as well as the empty one. A nominal height on a drawing is not evidence that the real robot can enter, align and engage throughout the cart family's service condition.

OMRON states that its OL-450S can handle many existing carts, trolleys and roll cages, and that integrators can adapt the lifting plate within defined guidelines. The company lists a 450 kg payload and a 108–308 mm lift range. Those figures describe that product; they do not make 108 mm a universal cart-clearance requirement or approve an arbitrary lifting surface.

Ask where the lifted mass is supported and how it remains stable. Include the cart, goods and relevant adapters in the carried-load calculation according to the manufacturer's rating definition. Confirm the permitted support points, contact geometry, load overhang and center-of-gravity envelope. A frame designed for support at its caster mounts may require a different assessment when supported by a central lifting plate.

Finally, demonstrate set-down in the intended location. The robot must be able to release the cart without leaving it obstructing access, resting incorrectly or unavailable for the next pickup. Include the required state of any cart brakes and positioning features. A successful lift in a clear demonstration area covers only one part of the delivery.

When Should You Tow the Existing Cart?

A tugger AMR or automated guided tugger is a candidate when the carrier is designed to be towed and its wheel arrangement suits the route. A regular replenishment loop may support several carts moving together. Point-to-point work may instead favor individual pickups. The right comparison includes train formation, connection work and delivery timing, not just a headline towing capacity.

Jtec's TugBot AGV product information describes automated towing for compatible tongue-and-hitch cart systems. That provides a concrete example of retaining towable carriers. The supplier's compatibility statement should still be translated into a documented list of approved carts, loads, couplings and operating conditions for the proposed installation.

Do not compare a towing mass directly with a lifting payload. In a towed arrangement, the cart's wheels support its weight and the vehicle must generate the required pulling and stopping behavior for the system. Wheel condition, rolling resistance, floor transitions, slopes and the coupling arrangement affect the engineering assessment.

Trace the entire train through corners and intersections. A robot body can pass an obstacle while a following cart takes a different path. Include the last cart, extended handles, coupling movement and recovery access in the route review. If reverse movement is needed, require the supplier to demonstrate the permitted maneuver for the actual combination rather than assuming that forward towing proves it.

Separate automated travel from automated connection. A quotation may automate the route while leaving hitching and unhitching to an operator. That can still be a valid design, but the remaining labor, access task and dispatch dependency belong in the business case.

Build a Cart Compatibility Record Before Asking for a Price

The most useful cart compatibility assessment groups carriers by relevant differences. Two carts with the same purchasing description may have different wheels, repairs, crossmembers or pickup features. Conversely, different upper racks may share a consistent transport interface. The family definition should follow what affects movement and handoff.

Create a record for each family with a revision identifier, drawings or measured dimensions, representative photographs and the permitted load configurations. Record condition variation as well as nominal geometry. The goal is to prevent a demonstration using one selected cart from becoming an unsupported promise about the entire installed population.

Then assign a provisional outcome: retain, retain with specified modifications, exclude pending repair, or replace. “Compatible after modification” should name the modification and its owner. “Compatible subject to survey” should identify the unanswered questions. Neither should be treated as unconditional compatibility in the purchase order.

Connect these records to the AMR site survey. A favorable cart assessment does not resolve a doorway, floor joint or handoff location that prevents the configured system from completing the route.

Why Cart Waiting Time Can Change the Preferred Arrangement

Count transport resources and load carriers separately. A facility with 40 carts does not automatically need 40 robots. However, converting every cart into a powered autonomous unit attaches automation to the carrier inventory, including the time those carriers are stationary. A separate robot can be shared only if the process actually allows it to disengage.

Consider this original planning illustration. The required flow is 12 complete cart circulation cycles per hour. Each cycle includes six minutes of transport-resource work across pickup, delivery and empty return, plus 18 minutes of stationary use at the destination. For simplicity, assume the same six minutes of robot work for both concepts; an actual study must measure their different handling times.

These are arithmetic lower bounds, not fleet recommendations. They exclude charging, congestion, failures, travel imbalance and service reserves. The integrated-cart result does not determine how many individual carriers must be converted to support the full circulation and inventory policy. The shared-robot result also requires enough compatible passive carts and usable set-down positions.

The example shows why dwell matters. At 18 minutes of destination use and 12 cycles per hour, the mean destination occupancy is 3.6 carts under the stated steady-flow assumptions. Designing parking around that mean alone would still ignore arrival bursts and operational restrictions. A shared transporter saves no waiting time if it cannot release its load because the receiving positions are full.

Use an AMR fleet sizing analysis for the final resource decision. Test the peak period and handoff conditions rather than turning the five-versus-two illustration into a purchasing ratio.

Automated Pickup Is Often the Real Compatibility Test

Automated cart pickup must work with the carrier state the production process can reliably provide. A perfectly parked cart is useful for initial setup, but it does not establish how the system behaves when an operator leaves a carrier within the permitted parking variation or a load changes the frame's position.

Specify the allowed arrival condition at each pickup point. Define how the system confirms cart identity, correct orientation, interface engagement and readiness to move. A transport request is not proof that the intended cart is physically present. If the interface requires a particular brake state or an unobstructed entry space, show how that condition is established.

Run the complete handoff in the evaluation: approach, identify, engage, confirm, move, place, release and acknowledge delivery. Record human assistance at any stage. Ask what happens if the cart is absent, incompatible, obstructed or only partly engaged. Recovery should preserve a clear account of the cart, load and unfinished task.

A consistent pickup interface can be more valuable than retaining every original cart dimension. A modest adapter applied across a sound cart family may simplify service and replacement. An elaborate one-off adapter on every damaged or inconsistent carrier can do the opposite. Compare that engineering effort with a standardized replacement family before deciding that reuse is cheaper.

Compare Cart Transport Automation Cost Across the Whole System

Cart transport automation cost includes more than the robot or autonomous wheel kit. Compare each offer against the same production service: the required moves, permitted load families, handoff conditions and operating hours. A cheaper package with manual connection, lower availability or extensive excluded site work is a different scope.

Separate expenditure into three groups. The first follows the number of powered units: drive systems, batteries, chargers and software where licensed per unit. The second follows the cart population: adapters, mechanical changes, identification features, repairs and replacement carriers. The third follows the application: stations, integration, installation, validation, training and operating support.

Ask which assumptions change the quotation. Does another cart family require a new adapter design or only another copy of an approved part? Can the supplier support a replacement cart from a different fabricator? Who owns the interface drawing? What happens to the price if the robot must wait during unloading? These questions make future expansion costs visible.

A KUKA NorthShore Care Supply case study provides a relevant maintenance example. It reports caster wear and alignment problems with the previous towing arrangement, followed by a change to robots that lift and carry the carts. This is one deployment's experience, not evidence that towing is generally inferior. Its procurement lesson is to include cart maintenance and transport-interface behavior in the comparison.

Keep labor assumptions equally specific. Time released from pushing carts may support other work without removing a paid shift or position. Count the transport, connection, exception and maintenance tasks that remain. Record the intended operational benefit separately from any staffing or financial claim.

Make the Pilot Answer a Reuse Decision

An AMR cart transport pilot should resolve a defined purchasing uncertainty. For example: can this retained cart family support reliable pickup and release without changing the operator's material-access process? That is more useful than a demonstration whose only objective is to move a cart once.

Select representative carts from the condition range allowed for service. Include approved load configurations and the route features relevant to the proposed arrangement. Exclude damaged equipment that should not be in service; a pilot should not normalize defects by requiring the robot to tolerate them.

Agree on the measurement definitions and acceptance criteria before the trial. Distinguish completed missions from first-attempt pickups, and successful deliveries from deliveries requiring assistance. Keep timing distributions and exception descriptions, not only averages. Trial duration and sample coverage should follow the risks and variability being evaluated; no universal number of runs proves suitability.

The AMR site acceptance test matrix can organize the evidence. The cart-specific output should be a released family definition, an interface drawing, a list of required modifications and a clear responsibility for maintaining compatibility after deployment.

#AMRCartTransport #CartRetrofit #AutonomousWheels #TuggerAMR #Intralogistics