End of APO

What are the differences? And to what extent can companies benefit by selecting a non-SAP planning tool? The SAP Advanced Planning and Optimization (APO) supply chain planning system is used across industries worldwide. It is expected that maintenance and support for the solution will be discontinued in 2025. By then at the latest, users will have to look for an alternative.

Of course, the cloud-based successor solution SAP Integrated Business Planning (IBP) comes into consideration. However, the new product only works in conjunction with S/4 Hana. Therefore, the end of life of APO will be associated with a major migration effort.* Many companies are undecided whether they want to make this effort for a product that is not yet established on the market but is used for business-critical processes.

So if you're thinking about looking around for a more mature solution, you'd better hurry. Just looking around and specifying the requirements takes a lot of time. This also applies to implementing the new solution. If there is no longer any support for this after a possible rollback to APO, a risk arises.

SAP Integrated Business Planning is intended to offer more performance and close the gap resulting from the end of life of APO. However, the successor solution does not replace all previous APO modules. For example, the Walldorf-based company has removed some central functions for detailed planning and order scheduling (Order Promising) and integrated them into S/4. Only the modules for material requirements planning and for procurement, production, sales and transportation of materials can be found in IBP in a revised form.

IBP and S/4

Users have to deal with the question of which functions of the existing APO platform they will find in S/4 and which in IBP in the future. In addition, the splitting of functional areas could have a detrimental effect on the coordination between different players within a company's supply chain management.

The distribution of features across different applications means fragmentation. However, the trend in the market is towards integrated end-to-end platforms. These not only provide a holistic view of the supply chain. They enable the analysis, planning and execution of processes within the same product family.

Looking beyond the horizon

So now, at the latest, it's worth thinking outside the box: What alternatives to IBP as an APO successor solution are conceivable? For example, Blue Yonder's Luminate technology could be considered. The originally German company, an AI specialist from Karlsruhe, was acquired in 2018 by JDA Software, the provider of supply chain management software.

In 2020, the company was renamed back to Blue Yonder. The renaming of JDA to Blue Yonder reflects the growing importance of cloud and AI solutions. JDA's previous solutions have long been recognized in the market as an alternative to APO due to their end-to-end planning, supply chain planning and execution integration, modeling capabilities or control towers. The AI functionalities now available offer additional added value.

The company develops patented machine learning algorithms and integrates them into corresponding applications. The result is holistic machine learning solutions. The biggest challenge is to integrate machine learning reliably and repeatably into business processes.

Other providers are not yet so far advanced here: they merely integrate machine learning algorithms trained in an existing system environment, the output of which can then be used in the process. This means they provide an AI platform on which the customer must develop his own machine learning application.

The special feature of the Luminate technology is that the algorithms already take effect at the business process level in the sense of advanced analytics. Thus, they not only enable very reliable forecasts, but also the prediction and resolution of disruptions or the adjustment of various parameters along the supply chain, among other features. At the same time, their generation can be scaled to the extent that they can also be implemented in the environment of large enterprise applications.

Machine Learning for SCM

In addition, the results of the machine learning evaluations can be better explained and interpreted due to the direct reference to the business process, which increases the acceptance and trust of the users in the technology. It is therefore crucial that the machine learning application reliably and repeatably supports the respective business process to the benefit of the business users.

The support consists in the automation of processes. Particularly in the case of repetitive decisions that have to be made at high frequency, the machine is superior to the human. It is faster, more precise and less prone to error.

However, not only ordering and scheduling processes can be automated. Disruptions along the supply chain can also be quickly identified and countermeasures automatically initiated. If, for example, a delivery of goods is stuck at sea, the system is able to decide on its own where the missing parts can be procured from as an alternative, taking into account transport costs and routes as well as time and other relevant factors.

The algorithms learn from current and historical data. They make predictions about the probability of an event occurring. All subsequent decisions and the chain of business processes that is set in motion on the basis of the forecast are based on these.

These can be the simple rerouting of a truck, but also an adjustment of production plans in factories with consequences for downstream logistics processes. The more data there is and the better it is maintained, the higher the data quality. This has a direct impact on the quality of forecasts and thus on decisions and business processes.

Data diversity and algorithms

What is the source of the large amounts of data that the algorithms work with? Not only internal data from the ERP system comes into consideration, but also data from external sources, so-called externalities. This includes information about the expected weather, upcoming holidays, vacations, or even local location data.

A good machine learning application should provide a large part of this external data at the same time. In addition, information provided by the customers themselves is also relevant. In the area of industrial production, this includes historical data on the utilization of machines, order quantities, delivery volumes, but also data on the demand for certain products.

If classic ERP users now want to use a third-party planning tool, the effort for connection and customizing should not be greater than when using IBP. Blue Yonder, for example, has integrated interfaces and adapters into its Luminate platform that allow master data from the SAP ERP system to be seamlessly transferred to its own planning solution.

This ensures seamless real-time data exchange between the two system worlds at all times. Users can cover classic ERP tasks such as merchandise management, purchasing, accounting, cost accounting or controlling via their SAP system.

Typical SCM tasks such as strategic and detailed planning as well as transport and warehouse management can then be handled via Luminate. In addition, the integrated control tower ensures optimum transparency across the entire supply chain. If SAP users still do not want to do without IBP, a hybrid solution is also conceivable. In this case, customers can use basic functions from IBP and flexibly add features from Luminate as Software as a Service.

*Updated on July 23, 2021: SAP was asked to clarify the statement as follows: "The new product works independently of the ERP version and is therefore also interesting for companies that have not yet migrated to S4/HANA."