Load handling devices (LHDs) are the physical point of contact between an automated storage and retrieval system and every unit load that enters or leaves storage. A telescopic fork on a crane, a shuttle extraction mechanism, a rotating gripper, or a lift-mounted conveyor table must perform identically on the thousandth cycle as it did during acceptance. Commissioning and acceptance of these devices is therefore not a formal approval ritual; it is the systematic collection of evidence that the device is mechanically sound, electrically consistent, control-compatible, and operationally safe within its defined working envelope. This article provides a practical engineering checklist for warehouse operators, maintenance teams, and controls engineers who must verify an LHD before placing it into service or after a major repair. It explains what to observe, what evidence to collect, how to interpret common anomalies, and where the boundaries of site acceptance begin and end. Always defer to site safety procedures, lockout requirements, OEM documentation, and the judgment of a competent engineer before any activity involving live machinery.
Scope of Load Handling Device Acceptance #
An LHD in the context of AS/RS, cranes, shuttles, and lifts typically includes telescopic fork mechanisms, shuttle extraction plates, chain-driven or belt-driven puller systems, rotating head attachments, and lift tables with integrated conveyor decks. Acceptance should be defined at the boundary where the LHD connects to the mast, carriage, shuttle frame, or vertical lift platform, and extends outward to the load-engagement surfaces. Any components inside that boundary, including wear pads, guide rollers, chains, belts, pulleys, drive motors, encoders, proximity sensors, and cable carriers, fall under the acceptance scope.
The acceptance process must distinguish between three distinct phases. The first phase is static verification, where the device is inspected and measured in a de-energized state. The second phase is dynamic verification under controlled motion, both with and without test loads. The third phase is integrated validation, where the LHD interacts with the rest of the storage system, including racking, transfer stations, and inventory control software. A common error is collapsing these phases into a single continuous run; this creates acceptance gaps where mechanical issues are hidden by electrical response, or where software fault-recovery masks positional drift.
Device Types and Their Specific Acceptance Boundaries #
- Telescopic forks – acceptance includes straightness over full extension, synchronized travel of fork stages, and load-bearing surface condition.
- Shuttle extractor plates – acceptance focuses on plate flatness, friction interface, and consistent engagement depth.
- Rotating LHDs – acceptance adds rotation axes, centering accuracy, and orientation feedback.
- Lift tables with conveyors – acceptance includes platform leveling, conveyor tracking, and load transfer synchronization.
- Gripper or clamp-style devices – acceptance covers clamping force, jaw parallelism, and grip-hold verification.
Pre-Commissioning Conditions and Documentation Review #
No LHD should be commissioned until the environment and groundwork are verified. The area around the machine must be clean and free of debris, and the machine must be electrically isolated in accordance with site lockout procedures. Confirm that all mechanical guards, cable trays, and service platforms are installed as designed. Measure ambient conditions, particularly temperature and humidity, and compare them against the operating envelope stated in the OEM documentation. Extreme cold can stiffen seals and alter belt tension; high humidity can affect sensor performance. Record the conditions at the time of testing, because this evidence becomes the baseline for diagnosing seasonal faults later.
Review the documentation package before powering anything. This includes general arrangement drawings, wiring schematics, PLC I/O maps, bill of materials, and any specific maintenance instructions for the LHD. Verify that the installed hardware matches the drawings. Changes made during installation, such as alternate sensor brands or revised cable routes, must be captured as marked-up redlines. Unrecorded deviations are a leading cause of bearing failures and positional faults that surface months after acceptance. The commissioning engineer should also confirm that the software revision in the controller is the version that the OEM intended for this particular LHD configuration.
Mechanical Interface Verification #
The integrity of the mechanical interface between the LHD and its carriage dictates the lifetime of the device. Begin by inspecting all mounting bolting and locating pins for correct torque and orientation. Looseness here does not always produce an obvious sound; it often appears as a gradual increase in positioning scatter. Check the alignment of the LHD base relative to the direction of travel using a validated measurement method. A laser tracker or dial indicator setup is appropriate, but the chosen method must be repeatable so that future checks produce comparable data.
For telescopic forks, verify the straightness of the fork rails over the full extension range. This is typically done at partial extension, then full extension, measuring the deviation from a reference line in both the horizontal and vertical planes. Note the value at each stage, not just at the end point. A fork that deflects linearly under its own weight is normal; a fork that exhibits a sudden step or kink at a particular extension point suggests a damaged rail or a failed bearing. Similarly, inspect the chain or belt tension. A slack chain segment can cause a stage to catch up abruptly after a delay, creating an audible snap and placing unnecessary stress on the synchronizing mechanism.
Wear pads and guide rollers should be measured, not visually admired. Use a feeler gauge or caliper to record clearances at multiple points along the travel path. The tactile feel of the device during manual jogging, if permitted by site procedure, is also valuable: the mechanism should move without binding or gritting. Any deformation, scoring, or uneven wear pattern on the slides should be photographed and logged. If any component is already at or near the OEM recommended wear limit during commissioning, it should be replaced before acceptance, because commissioning cycles are far lighter than production duty.
Load Interaction and Geometric Fit Checks #
The most important functional test of an LHD is its interaction with a real unit load. Theoretical calculations based on nominal pallet dimensions are insufficient. Select at least three representative loads that reflect the variance seen in production: one at nominal dimensions, one at lower bound dimensions, and one at upper bound dimensions. If the site handles damaged or deliberately warped pallets, include a classification of such loads in the test matrix. Each test load should have a known weight and a marked center of gravity. Use a load that is offset in one horizontal direction to verify the LHD behavior under asymmetric loading.
Check geometric fit in terms of fork entry clearance, vertical under-clearance, and side clearance. The fork should enter the pallet opening without contacting the pallet structure, and the pallet should sit fully flush on the fork surface. If using a shuttle extractor plate, confirm that the plate engages the pallet bottom without pushing the pallet sideways and that no load edge hangs over the plate by more than the design allowance. These measurements must be taken at multiple rack depths and positions within the aisle, because rack settlement and floor tolerance affect the relative position between the LHD and the stored load. A clearance check performed at only one rack depth is not sufficient evidence of acceptance.
Record the outcomes in a table that captures the load ID, measured clearances, and any contact events. Contact events observed during geometric fit checks are not acceptable as temporary anomalies; they will appear only more frequently as components wear. If a fork scrapes a pallet at the widest accepted deviation, then the acceptance verdict should note the restriction of service until either the load envelope is constrained or the LHD is adjusted.
Actuation and Motion Profile Testing #
Dynamic testing begins with unloaded cycles across the full travel range. Each cycle should include the extension, dwell, retraction, and, if applicable, rotation or lifting. Observe the speed profile through the controller’s monitoring interface where possible. The device should accelerate smoothly, reach a stable cruise velocity, and decelerate before the endpoint without hunting or oscillating. Jerk transitions that are too abrupt can cause the load to shift on the fork, even if the final position is accurate. Record the cycle time and compare it against the design target. A cycle time that is slower than specification may indicate excessive mechanical friction or a motor sizing problem.
After several unloaded cycles, repeat with a load at the lower end of the design weight range, then with a load near the maximum. Monitor the motor current draw during extension and retraction. A significant spike at the same travel position on every cycle indicates a mechanical constraint, such as a misaligned rail or a binding chain. Also monitor the difference between loaded and unloaded cycle times; a small difference is expected, but a large difference points to insufficient drive margin. The LHD must be able to hold its position when stopped with a full-rated load, without drifting, over a defined dwell period. Any drift after stop is a safety-critical finding that must be escalated.
Positioning accuracy and repeatability should be tested over a longer sequence, at least 20 to 50 cycles, so that thermal effects and cyclic variations are captured. Measure the final load placement position relative to the target using the storage cell or transfer station as reference. Record a run chart of the deviations. A consistent offset in one direction is an adjustment issue; a random scatter band may indicate encoder noise, backlash, or loosely coupled mechanical components.
E-Stop and Safe Stop Behavior #
While under load, and in accordance with site procedures and OEM instructions, verify the behavior of the device during a controlled stop command and during an emergency stop. The device must stop, hold the load, and remain in a state that permits safe recovery. The load must not slide, tip, or shed from the fork. After the E-stop is reset, the recovery sequence should be initiated only through the authorized procedure, and the device should resume in a controlled, predictable manner. Document the exact behavior and any operator intervention required. These observations become the reference for troubleshooting later faults.
Sensor, Feedback, and Signal Verification #
LHDs are only as intelligent as their feedback devices. A load-present sensor that flickers can cause a full system stoppage, while a home sensor that is too sensitive to contamination can cause false zero references. For each sensor associated with the LHD, verify the following: the mounting bracket is rigid and correctly adjusted, the sensing face is clean and unobstructed, the sensing range is within specification for the target material, and the switching state is stable with no intermittent chatter. For inductive sensors, confirm that the target material is the same grade assumed by the sensor. A sensor set to detect steel is unreliable with stainless steel or aluminum targets.
Encoder feedback is another critical evidence source. Check pulse counts over the full travel distance and compare against the theoretical count for the mechanical transmission ratio. Discrepancies reveal issues such as slick wheels, broken teeth, or loose coupling. If the system uses linear encoders, inspect the encoder tape or rail for damage, oil contamination, and secure adhesion. Also verify that the reading head gap is uniform along the entire travel length. These measurements are best performed at multiple speeds because some encoders exhibit signal loss at high frequency.
Cable carriers and service loops for the LHD-mounted components must be inspected throughout their full travel. A rigid cable strain-relief in the carrier may be flexed only at certain points, creating premature cable fatigue. Observe the carrier for smooth motion without buckling. Inspect the cables for kinks, chafing, or exposed conductor. If any suspect cable is found, it is prudent to replace it before acceptance, since a cable failure in a service loop is notoriously difficult to locate once hidden inside a moving carrier.
Load Cycle, Fault Injection, and Recovery Boundaries #
Commissioning should include a continuous multi-hour loaded cycling run to expose thermal weakening, intermittent sensor issues, and intermittent mechanical noises that do not appear during short tests. Recommend an endurance block that reflects worst-case duty, including frequent short cycles and occasional long cycles, rather than a rhythmic sequence that masks variation. During the endurance run, periodically inspect the device for temperature rise at the motors, gearboxes, and bearings using an infrared thermometer or thermal camera. A component that is consistently ten degrees hotter than its siblings is an early indicator of an alignment or lubrication problem.
Fault injection is a controlled method of testing recovery boundaries without necessarily waiting for natural failures. This is performed by qualified personnel following valid site procedures and written approvals, and it must never compromise safety devices. A typical scenario is to simulate a blocked sensor, a lost encoder pulse, or a pallet that is positioned poorly at the pick-up station. The objective is to verify that the control system detects the anomaly, enters a safe state, and can successfully recover after the fault condition is cleared. Document the time to detect, the time to recover, and whether any manual intervention, such as pushing a button or jogging a motor, was required.
Recovery boundaries define what the system can and cannot do after a fault. For example, an LHD may automatically re-home after a positional error, but it must not automatically retry a complete pick cycle if a load was only partially extracted. The acceptance record should state these boundaries explicitly so that operations and maintenance staff know which sequences are safe to run without supervision and which require review by a competent engineer. A fault-recovery test that ends with a load in an unexpected position is a failed acceptance, regardless of how many successful cycles preceded it.
Diagnostic Reference Table: Common LHD Anomalies and Evidence #
The following table summarizes common symptoms that appear during commissioning, the evidence that should be collected, and the most likely areas to investigate. It serves as a practical guide for documenting observations consistently.
| Symptom | Observed Evidence | Likely Area | Suggested Diagnostic Action |
|---|---|---|---|
| Incomplete fork extension on intermittent cycles | Cycle time varies; motor current spike at same position; small positional error in encoder counts | Chain slack, pulley bearing, soft limit threshold | Measure chain deflection; inspect pulley bearings; verify limit switch timing against motion profile |
| Load-present sensor flickers during extraction | Sensor output state changes at a specific extension distance; vibration present near sensor bracket | Sensor mounting, cable strain, target geometry | Run sensor with bracket removed; check cable carrier for flex; use oscilloscope to capture signal dropout |
| Grinding or scraping sound on retraction only | Sound occurs on loaded cycles, absent on unloaded; visible metal dust near slide rails | Wear pad contamination, rail edge burr | Visually inspect slide rails linearly; feel bearings by hand-jog if permitted; record noise pattern location |
| Positional drift over long cycling block | Output position errors grow slowly; encoder count at home differs after warm-up cycles | Thermal expansion of fork structure, chain stretch, coupling slip | Record position at fixed intervals during endurance run; measure coupling grub screws; compare cold-to-hot dimensions |
| Pallet sits skewed on fork after pickup | One corner of pallet is visibly higher or angled; geometric fit check shows unequal side clearances | Fork stage synchronization, wear pad unevenness, pallet geometry | Measure fork height at multiple points on each stage; check left-to-right wear pad clearance; verify pallet under-clearance |
| Rotation angle offset in rotating LHD | Load orientation error increases with each rotation; encoder feedback non-linear | Rotation gear backlash, encoder coupling, mechanical stop deformation | Rotate manually over several turns and record feedback; inspect flexible coupling for slack; check torque of gearbox mounting |
Any anomaly recorded in this table format must include the timestamp, the run number, the load weight, and the operating mode. This evidence allows a maintenance engineer to reproduce the condition without guessing.
Documentation, Handover, and Maintenance Implications #
Acceptance is complete only when the documentation tells the full story of what was tested, what was found, and what remains open. The final acceptance package should include the commissioning protocol, the evidence records from every test, the table of sensor and encoder checks, thermal measurements, deviation run charts, and a clear list of any deviation from the original scope. The handover to the operational team must include the values that define normal operation, such as expected cycle times, allowable positional deviation, acceptable sensor behavior, and operating temperature ranges of critical components. This baseline allows operators to recognize early symptoms of degradation rather than waiting for a full failure.
Maintenance implications of acceptance findings should be communicated explicitly. If the LHD passes acceptance, the regular maintenance schedule should reference the baseline measurements taken during commissioning. For example,
