Inside Microsoft Dynamics AX

The Microsoft Dynamics AX product team blog

A Shift to Effective Demand Forecasting With Microsoft Dynamics AX 2012 R3

A Shift to Effective Demand Forecasting With Microsoft Dynamics AX 2012 R3

  • Comments 4

Demand forecasters increasingly want systems that combine simplicity and agility, yet are powerful enough to handle a high volume of transactions. The forecasting systems in high demand are those that are intuitive, complement existing user knowledge and capabilities to do more with less, and are flexible enough to adapt to changing marketplace requirements.

Organizations require powerful tools for analysis – not simply more raw data – so their users will have the right information, delivered at the right time and in the right format. In today’s business environment where IT investments are routinely scrutinized, business applications must not only enhance user efficiencies and drive overall business improvement – there also must be a solid ROI associated with the investment.

A good tool will significantly improve a planner’s ability to successfully meet customers’ needs while minimizing inventory carrying costs, which is always important to manufacturers.

Demand Forecasting With Microsoft Dynamics AX 2012 R3

When we release Microsoft Dynamics AX 2012 R3, we are planning to include a lightweight, yet powerful demand forecasting tool that enables users to forecast demand based on historical data, adjust the forecasted demand, and import the values into Microsoft Dynamics AX forecast models. By design, Microsoft Dynamics AX utilizes the basic concepts of gathering the historical data, taking that data and generating a statistical forecast based on that demand, and then giving the user the ability to easily make changes to those forecasting recommendations in Microsoft Excel.

Microsoft was able to utilize several of its existing solutions in order to create new functionality that will handle the demand forecasting requirements of our customers. As a result of utilizing familiar Microsoft technology, demand forecasting will be a cost-effective solution. To generate the baseline forecast, we rely on the power of Microsoft SQL Server Analysis Services. To visualize the forecast, adjust the forecast, and view or create new KPIs, we rely on the extensive functionality of Microsoft Excel.

What’s the Value Proposition?

We’d like to share with you a few of the proposed demand forecasting capabilities slated to be in Microsoft Dynamics AX 2012 R3:

  • Works Seamlessly with Microsoft Products

Demand forecasting will leverage Microsoft Dynamics AX2012 R3 in such a way that there is no need for data migration and integration (as long as you keep your historical transactions in the Microsoft Dynamics AX 2012 R3 database).

  • Familiar Tools

Planners can use a familiar Microsoft Dynamics AX user interface to determine the historical data range and parameters of the algorithm applied. You can read more about the forecasting methods and the parameters available here and more about time series forecasting in SQL Server Analysis Services here.

The forecast itself is provided in Excel and can be easily manipulated by the planner. Equally important, users running Microsoft Dynamics AX 2012 R3 on SQL Server Standard Edition will not have to upgrade SQL Server in order to use demand forecasting. However, there are certain performance benefits for companies running SQL Server Enterprise Edition or SQL Server BI Edition. These advantages occur because processing data mining models run in parallel in SQL Server Enterprise/BI edition.

  • Flexibility In Processing

Microsoft Dynamics AX2012 R3 will offer the capability to perform forecast at any order penetration point in the supply chain. Demand forecasting for Microsoft Dynamics AX 2012 builds upon this functionality and offers the ability to forecast both dependent and independent demand.

So How Does It Work?

The key to the effectiveness of demand forecasting is in the way it works seamlessly with Microsoft Dynamics AX 2012 R3, Excel, and SQL Server Analysis Services. The demand forecast generation starts inside Microsoft Dynamics AX 2012 R3. This forecast is quantity based and can be generated at different levels: item allocation key, released product variant, warehouse, customer, and by geographical area. Historical transactional data from the Microsoft Dynamics AX database is gathered and populates a staging table, which is then fed into a SQL Server Analysis Services mining model.

Demand Forecasting flow

Inside the mining model, SQL Server Analysis Services time series forecasting methods are used to calculate a baseline forecast. The parameters for the forecasting methods are set in Microsoft Dynamics AX 2012 R3. Changes made to the demand forecast and historical data from previous iterations are also made available in this demand forecasting cube.

In order to view and modify the baseline forecast, Microsoft Excel is used to connect to the demand forecasting cube. Any manual adjustments made must be published back to the cube before they can be imported into Microsoft Dynamics AX 2012 R3. Notice at any time there is only one demand forecasting cube, hence only one version of the forecast exists. This will give planners confidence in knowing they can trust the system.

After the manual entries/adjustments to the forecast are made in the demand forecasting cube, the adjusted forecast is imported back into Microsoft Dynamics AX 2012 R3 and used in forecast plans. The planner can use the generated forecast with the Master Planning capabilities in Microsoft Dynamics AX. Now, the production planners (or purchasing department) can leverage this valuable information to do their jobs. The key to this demand forecasting process is that it is executed in an intuitive, user-friendly environment which utilizes the power of Excel, allowing the user to quickly and easily update the data.

This simple design approach allows the systems to interact seamlessly while allowing the user to work in a familiar environment:

  • Microsoft Dynamics AX – to determine product family, algorithm parameters and range of historical data
  • Microsoft Excel – for usage, sharing and adjustment of the forecast

Microsoft Dynamics AX: Delivering Excellence

Delivering excellence is not just about providing best of breed functionality. It’s also important to make sure that you meet the individual needs of users while doing so within a reasonable budget. The design approach that we’ve taken with the planned demand forecasting functionality in Microsoft Dynamics AX 2012 R3 not only provides great functionality, but also utilizes existing Microsoft platform components (SQL Server and Excel) that most customers already use. With this demand forecasting functionality users are able to leverage historical data available in Microsoft Dynamics AX 2012 R3. This data is then turned into forecasting recommendations by applying advanced forecasting algorithms available in SQL Server Analysis Services. Finally, users are able to view and manage all of this data through a user-friendly and familiar work interface with Microsoft Office Excel. This “One Microsoft” approach helps customers achieve a low Total Cost of Ownership (TCO) while also providing a common platform and common user interface that makes for a quicker and more successful user adoption.

Using the demand forecasting functionality in Microsoft Dynamics AX 2012 R3, demand forecasters will be able to work in a system that is powerful enough to handle intense planning requirements placed on them while trying to predict future demand needs. They will have a tool with a user-friendly interface that is agile enough to adapt to the ever-changing business environment they live in where demand variability is the norm. This is an environment where Demand Planners will succeed and the manufacturing organization as a whole will benefit – an environment where they meet and surpass customer needs while minimizing inventory carrying requirements.

Ready to dive in and learn more?

  1. Learn: for more information on Microsoft Dynamics for Manufacturing, visit our Manufacturing web page.
  2. Watch: view the “Preview: Demand Forecasting for Microsoft Dynamics AX 2012 R3” session from Convergence EMEA 2013, click here if you’ve already registered for Virtual Convergence. If you’re not registered, click here to sign in for free access to this and other sessions.
  3. Attend: join us for the Microsoft Dynamics AX Technical Conference, slated for February 3–5, 2014, in Bellevue, Washington. Click here for more info.

/Bill Moffett

Microsoft Dynamics AX 2012 R3 is a pre-release product under development.  All features and dates are preliminary based on current expectations and are subject to change without notice.  The information herein is for informational purpose only and represents Microsoft’s current view as of the date of this blog. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft.

Leave a Comment
  • Please add 1 and 3 and type the answer here:
  • Post
  • The Virtual Conference doesn't seem to have the Preview posted.

  • You have to register for virtual Convergence and then you will find the session videos for the concurrent sessions.

  • This seems like a winner already. one question on forecasting the 2nd level, using the family, or allocation key.

    Obviously the history of these items does not consist of sales order issues but of productoin order issues. I assume the software does not differentiate these and just looks at 'issues', just like the standard Safety stock journal in Master planning has done for more then 10 years... Correct?

    So the items that are sold as spare parts and ALSO are issued to production, will just get a forecast based on a hybrid history of direct sale and consumption into production, it does not matter. that means that a forecast, thus generated ,will have to be consumed by "All transactions", the new parameter in 2012.. It seems to be perfect. Please confirm.!

  • To address Evert's first question: for demand forecasting, unlike safety stock journal, we have build a mechanism that allows the users to specify what 'issues' they want to take into account for the forecast calculation. Globally or per item allocation key, users can specify if by 'issues' they mean only sales order lines, BOM lines, sales quotation, etc or a combination of the above.

    In the case of items that are sold as spare parts and also used in production, users will chose to build history based on sales order lines and BOM lines, hence creating what you call a hybrid history. Than, you are right, they will use the All transactions reduction.

Page 1 of 1 (4 items)