Shippeo, a leading provider of global shipment visibility, has achieved a significant advancement in its predictive shipment ETA performance, aided by investments in new AI/ML infrastructure and sophisticated algorithms. As a result, shippers and logistics service providers are seeing improved accuracy and reliability of estimated times of arrival (ETAs) of their shipments, thereby increasing the efficiency and predictability of their supply chains.
he recent challenges caused by major supply chain disruptions have reaffirmed the need for organisations to improve shipment traceability on a global scale. Although it has never been easier to track shipments travelling throughout the world in near real-time thanks to more affordable devices and better connectivity, the type of data that supply chain management values is changing.
Shippeo’s Chief Product Officer, Anand Medepalli, says: “The question supply chain managers are asking themselves is shifting from ‘where is my shipment’ at the macro level to a more granular ‘When will it arrive at the next stop?’ and ‘Are there any risks of a delay?’”. This means that the ability to predict shipment ETAs at a very granular level is increasingly becoming a must-have, critical capability for shippers.
This improved ETA brings a range of benefits to manufacturers. Shippeo customers report that having accurate ETA predictions for shipment arrival times improves team productivity at key delivery sites, while boosting KPI performance, increasing sales and, of course, raising customer satisfaction levels. At the same time, these manufacturers, many of which are operating globally, are decreasing stock-outs or production line halts by as much as 76%, as well as reducing late penalties by 25%, dwell times by 30% and reducing unexpected freight costs.
As with any algorithm, ETA accuracy depends heavily on the quality of data inputted, which is why Shippeo’s core R&D focus is to continually enhance the accuracy and reliability of the data feeding the ETA predictions. In its latest third release, in addition to data quality, the company has revamped its machine learning (ML) infrastructure, as well as the methodologies and models its data scientists and ML engineers use. The result is an impressive 32% improvement in ETA accuracy up to 48 hours before a scheduled delivery.
