This application was made to give and share within the industrial Rhodia family important KPIs to follow.
In addition to the KPIS which were more or less know by all the industrial manager, a new axis was put in place, the Value Stream (VS).
A Value Stream is defined by a consistent set of products manufactured on a given industrial asset.
Reporting coordinator in charge of this application is Célia Gonzalez (from I&P Stream)
460 users are splited in the world.
They are Value Stream manager, plant manager or industrial director.
There is no critical period, this dashboard is uploaded once a month, the first week-end after the 6th working day, which is generaly the second weekend of the month.
This application has been developped in 2007-2008.
The industrial project leader was Jean-Louis Tartarin and the IS project Leader was Sylvie Kramdi.
The two developpers were Olivier Morard (Keyrus) for the data modeling and Nathanaël Quinson (Keyrus) for the Bex query and the dashboard in the workbook.
List of application role + menu role and explanation if we have several applications role with specials rules.
| Role Code | Role Description | Explanation |
|---|---|---|
| ZR_RCS_CA_M20 | VS - Value Stream | Role for the menu |
| ZBI_RCS_VS_A02 | Value Stream Analysis - Transactional Data - End User role | Applicative role given access to the infoproviders |
| ZR_RCS_VS_P* | Value Stream Analysis - Access to ALL Values Stream | Confidentiality role. One role per Value Stream given access to the VS slice of the cube |
List of autorisation objects mandatory for the application.
| Authorization object | Explanation |
|---|---|
| C_VALSTR | The authorizations are based on the Value Stream info Object |
Fixed costs Rules: In order to have simple rules, we have chosen to have multiples steps (working DSO) but with only one transformation rule in each.
Level 0: OD_CNPFA, this DSO is loaded with a delta mode from ODS_CCA9 which is used in CBS and contains all the fixed costs by cost centers.
There are 2 filters in this delta extraction:
The currency type = 10 in order to extract only the records which have the company currency (ODS_CCA9 contains also the controlling area currency)
The fiscal year period between Januray 2009 to december 2500, just to initialize the first delta extraction from January 2009.
In the start routine of the transformation, we also ahev a rule to delete all the records which are not "direct production cost center". To do this, the C_CCDPROD attribute of the Cost center (C_COSTCTR) is tested and we keep only the 'X' value.
Level 1: OD_CNPFC is loaded by a full update process from OD_CNPFA with the following selections:
The version of the CCR (0VERSION = 000)
The actual posting (forecast, budget, etc... are excludined) (0VTYPE = 010)
The fiscal year period with the 12 months of Y-1 and the months of the current year. This selection is calculated automatically via a variable.
In the start routine of the update rule all the records which are not related to fixed costs are deleted. The selection is done on the costelement group (C_COSTELE__C_CELTHF1) with the following values:
Level 2: OD_CNPF2 is loaded by a full update process from OD_CNPFC but also with a full update on itself with the following selections:
From OD_CNPFC: Fiscal Year = Y and Y-1
From OD_CNPFC, in the start routine there is also a rule to distritube the actualized fixed costs (ZRCS-AC and ZRCS-ACAMO) bewteen the value stream using a key determined with the absorption groups (ZRCS-ABCNP and ZRCS-ABAMO).
We also delete all the records which do not have the Value Stream filled.
From OD_CNPF2, in the start routine, we are calculating the fixed costs for the multi Value Stream only and only for the actual data (selection on fixed cost group = ZRCS-AC and ZRCS-ACAMO).
Level 3: CUB_VS is loaded in a full data loading using the Fiscal year period as selection. In the start routine we still have a rule to delete records which not have VS filled.
Variable costs Rules: The variable costs are collected from different sources:
OD_FVMK is used as a master data where all the costing variant and vrsion are stored by period and company.
OD_CPVS1 is emptied during each data loading process.
The MATY KPI: This KPI is used in the dashboard and we will display two indicators:
OD_CPV2 is emptied during each data loading process and reloaded from OD_CPVS1 in a full update mode without selection because OD_CPVS1 contains only Y and Y-1.
Production quantity Rules:
The production data are loaded in 3 steps:
OTIF Supplier Rules:
The Supplier OTIF is a KPI to measure if our supplier are delivering our raw material on time with the good quantity.
This data loading is done from DBSDTR03 with a selection on the calendar date (between the first day or Y-1 and the last day of the previous month of the year)
Then it is a direct link for the shipment costs
OTIF Customer Rules:
The customer OTIF is already made in an Oracle table, here we just have to extract the content of the table and to load in one to one mode
Stock inventory Rules:
For the inventory, we have a dedicated DSO to store the agregation by Value Stream, because the number of records into ODS_IC01 is very very big.
OD_VSSTK is loaded in delta mode after each ODS_IC01 data loading.
Then the CUB_VS is loaded once a month in a full mode with selection
Forecast Accuracy Rules:
OEE Rules:
JNI Rules:
CCR Rules:
This industrial dashboard is dependant to:
There are two dedicated process chain to load the industrial dashboard during the weekend.
Today this dashboard is launched manually because we need to make sure that the weekly CO-PC process chain launched friday at 6:00 pm CET is successfully finished.
This dashboard is uploaded once a month, the first week-end after the 6th working day, which is generaly the second weekend of the month.
Due to the control of the successfull CO-PC chain, the dashboard process chains are launched manually.
The time for this process chain is never the same because each month, the selection is different.
Begining of January, we will extract 24 months (full Y and Y-1), so it is the longuest process chain, more than 24 hours.
At the opposite, in february it will be the fastest because only 13 months will be lauded (Jan for Y and full Y-1)
The End users are not "playing" with the BEx Analysis/Analyser but with a dashboard developped in an Excel workbook.
All the documentation is in this GDoc document:
All the queries used in the dashboard are documented in the following document:
The user workbooks are dashboard with a lot of formula which are described in the previous files.
No more broacast.
The VS dashboard was broiadcasted to a dedicated eRoom where the users was able to acces and didn't refresh the dasboard by themselve.
No evolutions are planned since the Industrial dashboard should be replaced partiially with the i-Mep project.