Knowing where which goods are in demand when, before they are even needed, would make the logistics business much easier. Predictive analytics can predict this with a high degree of probability and Anticipatory Logistics can put it into practice.
There is no doubt that knowing the future is an enormous competitive advantage. The proverbial aunt Emma with her Grocery shop already knew that. Her forecasting tools included the calendar, the weather forecast, the ability to listen to her customers and a good deal of knowledge of human nature. From this information, she was able to predict the shopping behavior of each of her customers with a high degree of certainty: For example, Aunt Emma knew when she had to order more grilled meat from the butcher than usual. Because if the weather would be good at the weekend, Mrs. Huber could get a visit from her 3 sons and their families. And they all loved barbecuing in Frau Huber's garden. Just like some other inhabitants of the small village. So the shopkeeper ordered more barbecue food than usual. It would be unforgivable if the Hubers or even the whole village found too little meat at Aunt Emma.
In this example, Aunt Emma used data from the past (customer knowledge) and the future (weather forecast), analyzed them and used her experience to make a forecast of future buying behavior. This simulation of the future triggered a logistics process that brought goods that had not yet been ordered close to their potential customers and were available more quickly when needed than with a purely fact-based approach. Even as the groceries grew into supermarkets and customers became more anonymous, the retail trade remained true to the model of Aunt Emma in principle. With the introduction of scanner cash registers, stationary retailers were able to systematically collect large quantities of historical sales data and evaluate them into aggregated purchase forecasts. This information is decisive in determining which goods and in what quantities are kept in stock in the warehouse centres.
Digitization opens up new data sources
Digitization has led to a sharp increase in the number of available data sources. Each corner is tracked, monitored and recorded. The individual purchasing and search behavior of consumers is measured - sometimes anonymously, but all the more meticulously. The growing number of devices connected to the Internet of Things provides an even greater amount of data. Algorithms use this information to calculate findings, which in turn are then available as new data: Statistical techniques such as data mining, predictive modeling or machine learning can be used to calculate statements about the future state of a system from all these data. This prediction method is called "predictive analytics" and illustrates possible system states in the future. The more data used as a basis, the more likely the forecasts become. The next step would be to derive concrete actions from the future scenario generated by predictive analytics, which would bring about a desired state. If this method is used in logistics, it is referred to as "Anticipatory Logistics" or "Predictive Logistics".
Beer and strawberry biscuits before the storm
What now sounds very theoretical, the US department store Wal Mart has pre-exercised in 2004: Even then, the US department store chain already had a wide variety of data on its approximately 100 million customers. Wal Mart used this data to forecast consumer behavior using computer models. When Hurricane Frances threatened the Atlantic coast of Florida, the retail chain made targeted use of this knowledge: The calculation models have shown that the demand for beer and strawberry pop tarts before a hurricane massively increases. So Wal Mart made sure that the stores in the affected regions had both products in sufficient quantities in good time before the natural disaster struck. This enabled Wal Mart to better meet demand than its competitors.
Amazon patentierte Antizipatory Shipping
Amazon is also involved in Anticipatory Logistics and has already filed a patent for "Anticipatory Package Shipping" in 2013. The Group intends to shorten delivery times considerably. In return, it delivers goods to a delivery warehouse that customers in this region have not yet ordered, but are highly likely to actually do so. For this forecast, Amazon can rely on various sources - such as the purchase and search history, notepads or wish lists and the time the mouse pointer stays on certain offers. Norbert Brandau, head of Amazon's logistics location in Winsen, Lower Saxony, tells the medium Die Deutsche Verkehrs Zeitung how far "foresighted shipping" has advanced at Amazon: "It's not about sending the article off before the customer orders it. Rather, we want to predict the demand and distribute the goods in the network in such a way that we can bring them to the customer quickly and economically when ordering.
Amazon does not yet operate an "Anticipatory Shopping" where an individual customer receives a package with things that the retailer believes are highly likely to want at exactly this point in time. However, some Amazon customers still receive packages that they have not ordered: in order to promote certain products, the online retailer sends free samples to very specific customers. Amazon determines who gets them on the basis of customer data. However, this new form of product sampling is more of an additional business for the online giant, because the respective companies will probably have to pay something to send their samples. But this has little to do with "Anticipatory Logistics".
There are many possible applications for Anticipatory Logistics
The method is not only capable of shortening delivery times and optimizing warehousing. For example, it can significantly simplify the entire supply chain management: live data from the supply chain, such as GPS position and the current range of transport vehicles, information on the condition of the goods being transported, weather data, traffic reports, etc. - they all flow into forecasting models that can be used to predict future scenarios with a certain degree of probability. If this scenario is undesirable - such as the late arrival of an urgently needed spare part - countermeasures can be taken automatically or manually (supply chain event management).
Another possible application is the combination with predictive maintenance: This allows a more or less precise prediction of when a wear part in a system needs to be replaced. If the spare parts logistician is included in the predictive maintenance process, he can use "Anticipatory Logistics" to deliver the necessary replacement exactly when it is needed. Simulation-based planning, as practised by "Anticipatory Logistics", is at any rate one of the major goals of digitization in all industries, emphasizes Thomas Reppahn, Head of Central Logistics Product and Process Management at Schenker Deutschland AG, to Logistik Aktuell. At the same time, the expert points out: "If we have reached the point where we can calculate and control entire supply chains in advance, then the digitization of business processes can be regarded as almost complete. But there is still a long way to go - for all companies in the world."
Conclusion: Anticipatory Logistics: Why delivery will take place before the order in the future
Logistics is an industry that depends on many contingencies. Digitalization now offers the opportunity to measure more and more of these influencing factors precisely and partly in real time. The data evaluated using predictive analytics provide future scenarios that will occur with a certain probability. With "Anticipatory Logistics", the actual occurrence of these scenarios can be facilitated or prevented - as desired. The possible applications of "Anticipatory Logistics" grow with the degree of digitization. However, technology is still at the very beginning.
Image source Cover picture: https://www.pexels.com/photo/silhouette-man-standing-on-road-against-sky-during-sunset-327430/
