How do I do Data Reconciliation?

How to enter data, start a job, check progress and download results

Mark de Geus avatar
Written by Mark de Geus
Updated over a week ago

Applies to: Digital Plant Applications

Users who can do this: Plant Metallurgist, Plant Manager

The Data Reconciliation screen can be accessed on the brains.app left navigation bar. It displays a Job History table with details of previously executed scenarios, or jobs.

The Reconciliation screen allows you to :

  • Start a New Job

  • Reuse Job

  • Download the generated report

Let's see how to run a data reconciliation job:

  1. Click on the New Job button on the top right corner of the screen.

  2. Alternatively, if you would like to configure a job that is similar to a previous one, you can click on the relevant arrow in the Reuse column of the Job History list. This will automatically populate the data entry tables with the previously used values which can be adapted as required.

3. Navigate to the Reconciliation config screen where you can configure the job type, date range as well as specific job parameters.

4. The job type refers to a set of configurations that define the problem to be solved and is based on a set of symbolic equations that define the relationships between states.
The job types shown are Global reconciliation(Components) is used for reconciling a process with multiple valuable products in addition to the normal reconciliation job(Simplified).

Note: A future version will allow the user to define their own job types via a job configuration interface. Until then you are welcome to contact us and we will configure them for you.

Global reconciliation focuses on the total heavy mineral and material reconciliation,
Global reconciliation (Simplified) does not include auxiliary variables such as rejection fraction.
Global reconciliation (Components) includes/adds valuable metal mineral components reconciliation. Hence the latter gives a higher resolution reconciliation on the valuable metals/minerals (components) contained in the product.

5. Select the desired job type and press NEXT.

6. Enter the date range that the job refers to. Any automatic data aggregations from the historian (if configured) will be looked up based on this date range, and this is also the date that will appear in report titles.

7. Enter the measurements and their uncertainty bounds in the table provided. You can fill in configured suggested values for a specific variable by clicking the FILL ROW button or select FILL ALL from the top right to fill suggested values for all the variables. If you started the job with the Reuse option, the values used in the referring job will be already filled.

8. Automatically filled values will appear in purple, and if the field is edited the values will appear black, allowing you to keep track of changes made.

9. Once you filled in all the values, click the Process Job button at the bottom of the table.

10. Upon job processing, you will be navigated to the Job History screen, with the new job as the top entry and showing a PROCESSING status.

You can expect a typical job to run for about two minutes, after which the status will either change to COMPLETED or ERROR. You can configure and start a new job without having to wait for the previous one to complete.

11. If the job completed successfully, a link to the results in several different formats is provided in the results column.

12. If a job failed, you can hover over the error status to see more details about the error.
Note: We are working on improving the specificity of error messages to help guide the user to identify possible solutions.

Keywords: reconciliation, measurement error

Further reading:

Did this answer your question?