Demonstrating Machine Learning in Meter Billing

Overview

The figure shows the overview of ML processing for outsorted billing documents in SAP S/4HANA Utilities.

ML utilizes an extensive amount of previous utilities principal and transaction data to make predictions. This data is saved in the customer's system. ML accesses ML-relevant data for outsorted billing documents by using training and apply datasets. This information forms the groundwork for establishing the release confidence value for particular billing documents. Moreover, this data is essential for training a predictive model in ISLM.

The ISLM is enabled to define, train and activate a predictive model for the predictive scenario UTI_BI_OUTSORTED.

The report REML_BILLING_OUTSRTD_RELPRED picks all the outsorted billing documents based on the user-defined selection criteria. It then calculates the release confidence value for each outsorted billing document in the selection, provided there is enough historical data. Subject to a defined threshold and the calculated release confidence, outsorted billing documents can be automatically released.

Flowchart showing the training, application, and display processes in SAP utilities. In the Train stage, ISLM Scenario and Model UTI_MR_IMPLAUSIBLE interfaces with CDS View Training and Customizing Timeframe. During Apply, reports determine release confidence and automate releases, utilizing Utilities Master and Transaction Data, connected to Machine Learning Processing Data. The Display phase uses the SAP Fiori App to resolve implausible reading results, supported by a CDS Model offering machine learning data.

The report results are stored in a database table and can be displayed in the application log:

billing document
release recommendation
indicator about whether the documents have been automatically released

The report needs the following prerequisites:

The training and prediction timeframe must have been maintained in customizing.
A trained and activated model version for the ISLM predictive scenario UTI_BI_OUTSORTED must be available.
If billing documents are to be released automatically, a release threshold of at least 0.5 must be provided on the selection screen.

The SAP Fiori app Resolve Outsorted Billing Documents is used to display the release confidence for each outsorted billing document, if the release confidence was correctly determined and the outsorted documents have not been released automatically.

The closer the release confidence value is to 100%, the greater the assurance of the related ML prediction that the document can be released.

ISLM Scenario and Model

The figure shows the ML model training based on historical data and the label determination for outsorted billing documents in ISLM.

Flowchart illustrating the handling of Implausible Reading Results Historical Data. Starts with Reading Result Unprocessed, marked as Result Not Predictable, not relevant for model training. Other data are either Reading Result Corrected or Released, considered Result Predictable and relevant for training. Criteria include Status Change (Implausible to Plausible or Released), Consumption Characteristics, and Attributes like Meter Reading, Meter, Register, and Installation. A decision point evaluates criteria, resulting in a Label of either No Release or Release Document.

The predictive scenario UTI_BI_OUTSORTED works as an SAP template that facilitates ML in processing outsorted billing documents. This scenario encompasses an algorithm, a training dataset, an apply dataset, and a target variable, all of which help in training the customer-oriented predictive model. The training and apply datasets incorporate the pertinent features and labeling logic at the model level for outsorted billing documents.

Initially, the predictive model is untrained, indicating that there is no existing model version. It is necessary for the predictive model to be trained and activated within the customer's system.

The predictive scenario is linked to the application using the ISLM and report REML_BILLING_OUTSRTD_RELPRED, so that the predictive results can be consumed by the end user in their own environment.

In addition to the agent’s decision to release or reverse a billing document, possible subsequent actions are taken into consideration to determine the expected label. This might be that the document was released but subsequently reversed due to a specific reason.

The result is a release confidence value, which can be used either to process the outsorted billing document manually (label No Release), or to automatically release or reverse it without further user interaction (label Release Document).

Training Period

The customizing table TEML_OUTBILPARAM provides the training and prediction timeframes for outsorted billing documents.

Timeline diagram showing training and prediction periods. The Training Start Date is 03/01/2017, and the Training End Date is 01/01/2019, spanning 24 months. The Prediction Period extends beyond the training end date to the current date, 03/01/2019, with a 2-month gap between training end and today, denoting the timeframe for model predictions.

In this example,

the oldest document (billing period start) considered for training dataset is dated 03/01/2017 which is 24 months earlier than today and
the last document considered for training dataset is dated 01/01/2019 which is 2 months earlier than today.

The timespan between 01/01/2019 and today 03/01/2019 indicates the offset timeframe, during which possible actions could be performed for released documents, meaning that the billing documents within this timeframe should not be used for training.

The logic for the prediction period (apply dataset) also considers the lower timeframe; documents with a billing period start that is earlier than the lower timeframe will not be predicted, all other documents will be predicted