4.1 KiB
Timeline of Events
| Date | Event | Status |
|---|---|---|
| Oct 23, 2024 | ALPLA reports not receiving accurate data from Henkel | Initial Problem Report |
| Oct 30, 2024 | Brandon sends evaluation of all ALPLA-fed lines | Pending ALPLA Review |
| Nov 1, 2024 | David Busch urges remote validation of remapped information | Urgent Request |
| Nov 5-12, 2024 | Discovery of data conversion issues between systems | Problem Identified |
| Nov 15, 2024 | Signal exchange validated; bottles remaining count active on lines 40,41,42 | Partial Resolution |
| Dec 4, 2024 | Next-next IDH information now received from SAP | Progress Update |
| Dec 17, 2024 | Lenin requests variables di1-di9 to be changed from DINT to SINT | New Requirement |
| Jan 10, 2025 | Status update requested by Steve McEvoy | Pending Response |
Data Conversion Requirements
-
Original Data Flow:
- Input: String (from SAP)
- Intermediate: Three DINTs (Henkel's PLX module)
- Current: Two words per DINT (PLX conversion)
- Required Output: Original String format
-
Specific Tasks:
- Convert variables di1-di9 from DINT to SINT
- Implement changes in Q1 PLCs
- Update supervisory system PLC
- Validate data conversion accuracy
System Modifications
- PLC Changes:
- Must be performed in Italy
- Affects all Q1 PLCs
- Requires supervisory system PLC updates
Validation Requirements
- Verify correct string reconstruction from DINT values
- Ensure SAP data matches exactly after conversion
- Confirm proper data type handling for bottle count information
- Validate message mapping matches provided documentation
Critical Points
- All modifications to supervisory PLC must be done in Italy
- Changes need to maintain data integrity through multiple conversions
- System must handle both current and next-next order information
- Data type conversions must be consistent across all lines
!
During testing, we found that some values were being read correctly but in the wrong data type. Data from SAP goes through multiple conversions between Henkel and Vetro, as outlined below. On the left side of the PLX communication module, Henkel’s data is in one format, and on the right side, it’s Alpla's. The process is as follows:
!
SAP sends data as a ==string==. Henkel converts that string into three DINTs for the PLX communication module. The PLX then converts each DINT into two words. ==Vetro must then convert this back to the original string value to match the exact data SAP initially sent.==
During the testing, we observed that when comparing certain data points, Vetro was reading the correct DINT values. However, to retrieve the original string value, an additional conversion is needed on their end. On Line 41, all values were read back as string values, so every IDH, PO #, etc., matched exactly as Henkel receives them from SAP. On other lines, like Line 40, they received the correct values, and the DINTs matched those being sent. However, some data points still need to be converted back into a string to be fully usable. @Robert Nation, I think that the logic you created for line 40 with the new strapping table can be implemented across the OSBB fed lines to give Alpla a more accurate running bottle count now. This is assuming the tank dimensions are the same across the lines. Once completed, Vetro will need to verify that their message map matches the one provided to them and this should be resolved.
MAP L20