Predictive Analytics at DB Schenker
In logistics, we are faced with questions which require us to see into the future on a daily basis: How much freight will arrive at the terminal in four weeks time? Will ocean capacities are getting tight? Which customer may be abandoning us soon? When answering these questions we are increasingly relying on data-based evidence rather than gut feeling. To be precise, we are using predictive analytics. It has its origin in statistics and is used as a method within the field of machine learning which in turn is part of artificial intelligence.
How do we get these forecasts? We are getting them by evaluating large amounts of historical data which are available either within DB Schenker or from external sources, a mathematical model can be developed that is capable of identifying patterns in the data. The model can then be used to predict future events based on these patterns – so that we can develop solutions resulting in greater efficiency and lower costs.
Looking at the example of land transport, we have developed a tool which allows us to predict the incoming and outgoing freight volumes with up to 95 percent accuracy that will need to be handled in the coming days and weeks. Or looking at Supply Chain Risk Management: By enriching consignment tracking & tracing information with data about traffic volumes. the effects of weather conditions, waiting times at border crossings or strikes in ports, we can compute risk forecasts. Other applications of predictive analytics and artificial intelligence include market and capacity developments in ocean freight as well as an early warning system to identify customers who might be considering going elsewhere. Based on these predictions, we can adjust our capacities, optimize our tour planning and proactively take countermeasures.
Even today, algorithms will often produce better solutions, and quicker than humans can. It is now possible to make decisions that are based not merely on gut feeling but backed by hard data as well. Given the complexity and volatility of the markets, this is a definite advantage.
However, predictive analytics is not intended to replace human expertise at this stage but to supplement it. This will allow employees to focus on matters that are more mentally challenging than many of the run-of-the-mill tasks that still take up a lot of time at present - and of course to focus more on our customers.
Managing Director and Director Sales OEM Europe at Boeing Distribution
5 年Markus, supportive. I believe predictive analytics will become even more important if looking at environmentally sustainable business models. Predictive analytics will support removing waste (in its true sense) in processes.?
Digital Transformation Excellence ?? Lecturer & Writer ??Mentor at Plug & Play Tech Center München ????Keynote Speaker??Passionate runner??♀?Art Connaisseur ????
5 年This is indeed a very interesting point of view . Neverthenless I am asking myself how secure is your digital ecosystem? How sure can you be in your prediction? What if....#digitalvaluecreator #machinelearning?@klausstolper #cybersecurity
Financial Controller | Strategic Support | Commercial Finance
5 年Markus Sontheimern a very good article indeed. Two points I would like to share, that the accuracy and usability of results will depend greatly on the level of data analysis and the quality of assumptions.
Vice President Advisory, Marketing & Sales at Opticoms
5 年Markus Sontheimer interesting read indeed. Predictions always depends on two factors in my view: a) how good is the quality of the available data? b) how quickly can I get insights out of it? Available data certainly has to include live data as well. In our world a lot of data gets outdated rather quickly, so that needs to be accounted for. Also the bigger the (usefull) data set is the better for the prediction. That usually results in a problem with the speed. But luckily there are solutions available with can ingest, process, analyze, and visualize billions of data points in real-time giving your business instant insights.
Co-Founder @Nalia
5 年Really nice article, thanks for sharing. 2 small observations on my side: - predictive analytics is a powerful tool when we use historical data to build patterns and predict future behaviors, it’s even more powerful when we enrich data first with live & public information, such as wether in some cases (you already talked about it), but also stock prices or web scrapping in others. - as you said in your conclusion, predictive analytics won't replace humans, and could be considered as a powerful help to the decision-making tool. After a monitoring phase and a user validation, it could become a decision-making tool, transforming redundant decision processes into automated and powerful processes?