Innovation on SAP HANA Platform – Intelligent Stock Balance System Across Stores
One of the best parts of my job is continually discovering new instances of how SAP HANA is being employed to solve real-world problems. I’m always impressed with the variety of SAP HANA use cases. Thanks for Intel’s help to collect much more details from customers.
For example, Hong Xiang Yi Xin Tang (Hong Xiang) – a gigantic enterprise consisting of more than 2,600 pharmacies and one of China’s largest and most rapidly growing retailers – paired SAP HANA with Intel’s latest processor technology to solve the highly complex challenge of precise cross-store merchandise re-allocation.
As you might expect, a company like Hong Xiang has an absolutely massive inventory containing millions of items. Long ago, they established a well-planned system of standardized and centralized warehouse-driven replenishment procedures. But there was one glaring gap in their centrally controlled arrangement: specifically, it was quite common for an individual store to suddenly encounter a serious inventory shortage, including critical medications. This could happen for all sorts of reasons, such as promotions, geographic and location factors, and regional incidents such as illness outbreaks.
Store managers who were blindsided by a shortfall typically sought help by hastily reaching out to their colleagues in other retail locations. This resulted in a tedious, time-consuming, and error-prone process of manually creating spreadsheets, making phone calls to other stores, and then pleading for stock replenishment. Everyone – the company, the stores, and worst of all, the customers – lost out when this happened.
Hong Xiang and the SAP Active Global Support organization teamed up to create a collection of comprehensive, sophisticated inventory optimization algorithms that served as the foundation for a new cross-store merchandise replenishment solution. The new application was developed using SAP HANA, which has been optimized to fully exploit the advanced vector extensions (AVX) offered by the Intel® Xeon® processor family. For this system, it requires fast data analytic power to calculate current stock for each store, predict the demand changes and estimate the out-of-stock situation, and the optimize the delivery options. This needs prediction and optimization based on the store inventory data. Intel® Xeon® processor family offers the SAP HANA multiple CPU multiple core architecture to run those complex calculation in an extremely parallelized way.
Every morning, the system kicks off a far-reaching series of predictive analytic calculations to identify potential overstock or product shortage situations. This demands powering through billions of computations that balance user-requested stock levels with historical sales data that includes all stores, and is particularly attuned to location, cost, and gross profit considerations.
Even though there’s an immense quantity of permutations to get through, the SAP/Intel tandem finishes the job in near real-time and enables the application to determine the optimal product allocation distribution for each store that day. Store manager instantly get alerts about imminent shortfall or out-of-stock situations, which lets them manage by exception. Once they’ve learned about potential trouble, a single click launches the solution’s allocation optimization module, where store-specific settings related to priority, inventory levels, and other fine-tuning details may be adjusted. The system also applies a rich set of algorithms to suggest the most advantageous delivery mode (e.g. bicycle, motorcycle, or van) and route. Regional managers have oversight options for their own group of variables.
It’s no exaggeration to point out that the joint SAP HANA/Intel solution has transformed Hong Xiang’s daily operations – for individual stores as well as the whole company. First, every store manager now has access to vital information around the clock, and is able to make better decisions and more precise sales forecasts in real time. This improved accuracy and visibility also means that there is no longer a need to pre-emptively counter unexpected inventory outages by stockpiling excess merchandise. Meanwhile, since inventory is more rationally allocated across the far-flung retail network, Hong Xiang’s merchandise and transportation costs are lower, leading to improved profitability. Emergency events such as illness outbreaks or more mundane issues like normal merchandise imbalances can be handled gracefully and with well-informed coordination across all stores.
Best of all, Hong Xiang’s customers are able to realize the advantages of better service and more predictable merchandise availability. Confidently anticipating that the products they require will be obtainable at a nearby location helps strengthen their brand loyalty and is a fundamental pillar as Hong Xiang continues to grow.