Key takeaways
- Select the AMR from the complete loaded condition, including mold, carrier, attachments, fluids, and off-center mass.
- Approve the route from individual wheel loads and slab details, not payload rating or average pounds per square foot.
- Use mechanical locating features and positive load restraint instead of expecting navigation accuracy to align the mold alone.
- Treat every crane handoff as a controlled state with one authority, an exclusion zone, and verified equipment positions.
- Prove the hardest mold, weakest floor segment, and least forgiving dock during a representative pilot.
Can an AMR safely carry an injection mold?
Yes, an autonomous mobile robot can move heavy molds among storage, maintenance, and molding machines, but only when the application is engineered as a tooling-transfer system. Rated payload is merely the first gate. The real decision turns on wheel loads, center of gravity, braking behavior, docking tolerances, load retention, and the handoff to cranes or machine-side equipment.
A suitable AMR must keep the mold stable through turns and stops, travel only over a structurally approved route, and dock closely enough for a separate mechanical interface to complete alignment. It also needs a defined control sequence that prevents the robot, crane, press, lift table, or personnel from entering the same hazardous movement state.
For most plants, the best first step is not a catalog comparison. It is a route and interface study using actual mold weights, dimensions, pickup points, storage geometry, press heights, floor records, traffic patterns, and crane procedures. That evidence shows if repetitive transport automation is practical and what equipment architecture the plant truly needs.
What load case should engineering use?
Begin with the heaviest credible operating condition, not the mold’s nominal weight. Add the AMR, top module, transfer plate, adapter frame, hoses, couplers, retained oil, and any cart or skid. Confirm the mold’s center of gravity in both axes, since an off-center tool can overload one wheel or reduce stability well before the total reaches the platform’s advertised capacity.
Consider a hypothetical 24,000-pound mold on a 4,000-pound robot and carrier. The combined static mass is 28,000 pounds, and a simple four-wheel average is 7,000 pounds per wheel. That arithmetic is only a screening value. Acceleration, braking, floor slope, caster geometry, frame stiffness, turns, and load eccentricity can shift substantially more force onto one contact point.
The engineering file should therefore include individual wheel reactions for the worst travel and stopped conditions, allowable center-of-gravity envelope, maximum grade, cross-slope limit, turning speed, emergency-stop case, and required safety margin. Check that the structure and retention hardware remain acceptable after a single drive-wheel fault, abrupt stop, or small positioning error at the dock.
Can the floor carry concentrated wheel loads?
A heavy AMR does not apply its load evenly across the aisle. Each wheel presses through a small contact patch, creating concentrated forces that may govern slab bending, punching, joint distress, and local surface damage. A floor that has carried ordinary forklift traffic is not automatically approved for a different wheel layout, tire hardness, axle spacing, or repeated route.
The American Concrete Institute’s ACI 360R-10 guide treats vehicular loads, concentrated loads, joints, and load-transfer mechanisms as distinct design topics. That is the right mindset for mold transport. Review slab thickness, concrete strength, reinforcement, subbase condition, joint details, embedded utilities, trenches, pits, patches, rack anchors, and the distance between the wheel path and slab edges.
Give a structural engineer the actual wheel reactions and contact areas, not just total vehicle weight or pounds per square foot. Survey the full route for rocking joints, spalls, cracks, drains, steel plates, thresholds, and previous repairs. OSHA’s warehousing guidance also calls for cracks to be filled and transitions smoothed, which matters because a sharp lip can produce impact loading, wheel slip, vibration, and navigation error at the same instant.
How precise must docking be?
Required docking precision comes from the receiving interface. A maintenance table with broad lead-in guides may tolerate more error than a press-side rail, powered roller bed, or narrow mold shoe. Split the tolerance budget among AMR repeatability, floor variation, frame deflection, lift-height accuracy, mold variation, and wear in the receiving fixture.
Do not ask navigation alone to perform final alignment. Use tapered guides, hardened stops, locating pins, side rollers, or another mechanical capture feature to convert a near dock into a repeatable transfer position. Presence sensors should verify the correct mold, full insertion, load support, restraint engagement, and a clear transfer path before powered motion begins.
Test docking with dirty sensors, worn wheels, changing light, reflective machinery, and the full range of mold mass and center of gravity. Approach speed should fall in stages near the interface. Record position and height results over many cycles, including recovery after an obstructed approach, rather than accepting one successful commercial robot demo as proof.
Which lifting interface fits the mold flow?
The top module must match how the mold is stored, serviced, and presented to the press. A lift deck can change height between stations. Rollers or chains can transfer onto compatible beds. A retained carrier can remain beneath the mold, while a low-profile platform may travel under a purpose-built stand and raise the load clear.
No interface is universally best. Roller transfers need positive stops so gravity or vibration cannot release the tool. Hydraulic lifts need controlled descent and protection against hose failure. Screw or electromechanical lifts need overload detection and a safe response to position disagreement. Every design needs a method that keeps the mold restrained through travel, power loss, docking, and transfer.
Standardize the receiving geometry where practical, but do not conceal press-to-press differences behind a nominal common height. Measure each storage bay, maintenance bench, preheat station, and molding machine. Adapter plates can absorb variation, provided they are keyed, rated, inspectable, and impossible to install in an unsafe orientation.
What belongs in the safety and control design?
ISO 3691-4:2023 specifies safety requirements and verification for driverless industrial trucks and notes that operating-zone conditions significantly affect safe operation. In the United States, ANSI/A3 R15.08-2-2023 addresses industrial mobile robot systems and application integration. Use the applicable standards as inputs to a site-specific risk assessment covering normal production, foreseeable misuse, recovery, maintenance, and abnormal stops.
The molding cell adds its own hazards. ANSI/PLASTICS B151.1-2017 addresses known hazards to people working on or beside injection molding machines, including the mold and clamp areas. Interlocks should establish clear states such as machine unavailable, dock ready, robot secured, transfer enabled, load confirmed, and departure permitted. Conflicting or stale signals must produce a safe stop.
Mold changes and maintenance can expose workers to electrical, hydraulic, pneumatic, thermal, mechanical, and stored energy. OSHA 29 CFR 1910.147 requires energy-control procedures when unexpected startup or energy release could injure employees. An AMR command, software pause, or machine stop button is not a substitute for an energy-isolating device when lockout and tagout apply.
How do you prove the system before committing?
A useful robot pilot program reproduces the hardest work, not the cleanest demonstration route. Use the heaviest and most eccentric approved molds. Cross the weakest joints, execute the tightest turn, dock at every interface type, test traffic conflicts, and rehearse blocked routes, low battery, sensor contamination, network loss, failed handshakes, and manual recovery.
Service Robot Co. approaches this work as an OEM-neutral, full-service commercial robot integrator for U.S. businesses. The team can compare equipment across manufacturers, perform site assessment mapping, arrange financing, complete robot deployment and integration, train operators, and provide service through a nationwide U.S. engineer network. That gives the plant one vendor for the full operating lifecycle.
Commercial robot rental, AMR rental, and robot leasing for business can support a staged deployment when the commercial structure fits the plant. Ask how monthly payment programs, maintenance included terms, on-site dispatch, remote triage, go-live support, spare coverage, and end-of-term choices are documented. The commercial model should support engineering discipline, never replace it.
- Approve the full loaded mass, center-of-gravity envelope, and individual worst-case wheel reactions.
- Obtain written structural acceptance for every route segment, dock, joint crossing, pit edge, and maintenance position.
- Set measurable limits for approach position, lift height, final mechanical location, cycle time, and failed-dock recovery.
- Verify positive load retention during travel, emergency stopping, power loss, and transfer interruption.
- Prove crane-zone ownership, warning methods, exclusion controls, and recovery after a communication fault.
- Run repeated trials with representative debris, lighting, traffic, wheel wear, and battery state.
- Train authorized operators and maintenance staff on normal use, manual recovery, inspection, and energy control.
- Record acceptance evidence by mold family and station rather than approving the fleet from one favorable test.