AI & Predictive Analytics in Freight Forwarding

Predictive analytics is described as a range of sophisticated techniques and tools to analyze and interpret data to extract insights and foresights and provide actionable intelligence beyond what is offered by traditional business intelligence (BI) methods. Together, AI and predictive analytics are transforming freight forwarding and the supply chain in many ways. Some key examples include helping to streamline operations, improve efficiency, and reduce costs. 

For example, AI and predictive analytics can help optimize routes, forecast delivery times, and manage risks like traffic congestion and weather disruption. They can also help anticipate demand fluctuation, enabling better resource allocation and inventory management. They enable better decision making by providing real-time insights and allowing proactive adjustments to processes and schedules across the supply chain. And for shippers, AI and predictive analytics can ensure faster and more reliable shipping services.

Large volumes of logistics data must be analyzed. To do this, AI relies on algorithms that can learn from patterns in the data. Equipped with AI algorithms, the machine can now apply the learned knowledge to undertake tasks that would normally require human intelligence. In addition, the algorithms can utilize experience to improve over time; the more data an algorithm processes, the more accurate its predictions and decisions become. 

Predictive analytics leverages a range of methodologies, such as statistical analysis, predictive modeling, machine learning (ML), and data mining. Today, rather than merely having access to large volumes of data in the supply chain, logistics professionals can intelligently use and analyze this data for strategic advantage, which is essential to answer the increasingly complex and specific questions asked by businesses today.

Among many analytics solutions, DHL offers Smart ETA for port-to-port shipments on ocean voyages. It is included free of charge and all ETA information is accessible to DHL Global Forwarding customers on the myDHLi digital platform .

At its core, this predictive analytics solution has an automated ETA forecasting engine, providing a single point of truth. Near-real-time prediction is achieved by including frequent ETA updates during transit, and our aim is always to minimize the gap between the estimated and the actual time of arrival.

Our first freight forwarding forecast is provided when the vessel departs – at this point, the Smart ETA solution calculates the arrival time based on historical carrier schedules for this vessel on the same route. As the journey continues, the Smart ETA solution uses GPS positioning data from an automatic identification system (AIS) to repeatedly adjust the arrival time. It also uses harmonization methodology to reduce reliance on conflicting and inaccurate ETA updates, improving overall accuracy.

On our current robust foundation, we run a supervised machine learning operations model in a constant and automated way. This allows us to experiment and improve our Smart ETA predictions in the future and it will provide actionable insights from which to make informed freight forwarding decisions. Also, we ensure this model is becoming more adaptive to global changes – learning how to quickly identify and adjust to emerging and exceptional crises that are likely to have significant impact on the duration of ocean crossings. 

With focus on continuous improvement, we are utilizing experimental science to calculate how close we come to reliably predicting each ETA in the supply chain. We are also working to enhance not just the last leg of each ocean journey but also ETAs through the entire shipment lifecycle. Ultimately, we intend to achieve a full predictive approach for end-to-end forecasting with multiple accurate ETAs, including port stops along the way, and the estimated likelihood of making scheduled transit connections.

Another freight forwarding development on the near horizon is our use of generative artificial intelligence (Gen AI) with the Smart ETA solution. This will allow interfacing with our vast knowledge base to obtain new insights. For example, a junior analyst would be able to ask our chatbot to “Build a chart comparing shipments” and this would be swiftly created. No need for a subject-matter expert, no need for advanced data training, and now your company can spend less time on shipping and more time on core business activities.