A YASKAWA PL80 pallet exchange fault can look like a robot position problem while the cell is waiting for a pallet-ready or layer-state condition. Preserve the first missing state and the physical pallet position before editing the job. A robot at the expected approach point does not prove that the next pallet is detected or supported.
Yaskawa lists PL80 as a palletizing robot with a YRC1000 controller context. The product page does not define the installed pallet sensors, conveyor interface, gripper or recovery program. Use the cell electrical map, pallet drawing and actual job state for diagnosis.
Freeze the first exchange event
Record the current layer, pallet identifier, exchange request, pallet-present input, safety or cell-ready state and the next program step. Note whether the old pallet left its position, whether the new pallet reached its datum and whether the robot had a part in the tool. Keep the timing order of signals and physical movement.
- Pallet datum and guide condition.
- Sensor target, input name and logic state.
- Conveyor or transfer handshake and timeout branch.
- Retained layer and part state after interruption.
Inspect pallet datum and support first
After the load is controlled and the cell is in the required safe condition, compare the pallet position with the approved datum. Check guides, stops, debris, support surfaces and the part that may have shifted. A sensor can report presence while the pallet is not seated against its reference. Record physical gaps and direction before changing a sensor target.
If a pallet is skewed, a release or approach position can be wrong without any robot data change. Separate pallet handling work from a robot frame correction. The PL80 model name does not prove that the transfer mechanism or guide is installed as shown in a product illustration.
Trace the handshake from request to ready
Use the installed signal map to list the request, transfer busy state, pallet-present input, clamp or stop confirmation and robot permission. Record which device produces and consumes each state. A screen message such as “waiting for pallet” may be a combined condition. Find the first false or missing state before changing a timeout.
Compare the signal timeline with the actual transfer movement. A successful conveyor movement with no input suggests sensor, wiring or mapping investigation; a present input with no movement suggests the transfer or mechanical branch. Keep PLC filtering, inversion and pulse duration in the record.
| Observation | Compare next | Not established |
|---|---|---|
| Request never issued | Job and layer logic | Transfer hardware failed |
| Transfer moves; no presence input | Sensor, target and map | Pallet is absent |
| Presence input; pallet not seated | Datum and guide | Robot frame is wrong |
| Ready state; recovery waits | Retained layer and part state | Timeout is too short |
Separate layer state from pallet identity
A new pallet may be physically present while the job retains the previous pallet or layer number. Save the identifier selected by the cell and the one expected by the robot job. Do not reset a layer counter merely to make the next motion continue. Confirm how the application records a completed layer and a partially placed layer before resuming.
If a part remains in the gripper or on the pallet, make that physical state part of the recovery decision. A repeated release command can create a dropped part, and a skipped layer can create a later stack error. Keep the operator confirmation and the sensor evidence separate.
Compare the pallet exchange record with the stack pattern after the next approved cycle. An exchange can be successful while the retained layer number is wrong, causing a later offset that appears unrelated to the pallet. Save the first correct placement and the first incorrect layer as separate evidence.
Where a pallet detector is shared with another cell function, identify the interlock and its filtering. A signal may be true for transfer permission but not for a seated pallet. Do not repurpose a convenient input or bypass a wait without documenting the cell design and recovery review.
Record the pallet detection input itself. Conveyor motor motion or a timer is indirect evidence. If the installed sensor is not visible in the job, use the electrical drawing and PLC diagnostic to identify the input. A green transfer indicator may mean only that the request was accepted.
When the PL80 is waiting at a pallet position, record whether the robot is holding a part and whether the gripper confirmation remains true. A recovery move made with an unknown part state can change the stack even if the exchange itself later succeeds. Keep this operator decision in the handover.
Repair the confirmed transfer or integration boundary
A damaged sensor or cable needs its matched replacement and signal verification. A pallet guide problem needs mechanical correction. A mapping or retained-state issue needs a backed-up integration correction. Only independent motion evidence should lead to PL80 robot-component work. Record the owner and verification for each action.
Verify exchange, layer and restart together
Run representative exchanges with the agreed pallet, guide, gripper and part condition. Verify old-pallet removal, new-pallet datum, handshake, layer selection, placement and the permitted interruption recovery. Record repeat count and any part movement. A no-load conveyor cycle does not validate the robot job or stack result.
Handover should list the physical datum, signal names, layer and pallet identifiers, changed component or logic and remaining limitation. Keep the pre-change backup and the final approved version. This makes a later stack error traceable to the exchange record instead of an unexplained reteach.
Palletizing robot inspection and repair
Contact Suntech with the PL80 exchange record