Modulai applied a set of machine learning techniques to predict the risk of shipment failure in temperature-sensitive pharmaceutical goods. The models, trained on hundreds of thousands of historical missions, can optimize packaging choices, transport routes, and serve as the basis for a real-time warning system for operators.
Stats
20K+
Users worldwide
100K+
Historical missions for training
30+
Years of client experience
Challenge
The client is one of the world's leading providers of solutions for tracking sensitive pharmaceutical goods during shipment and throughout the supply chain. Each shipment is tracked minute-by-minute by a temperature logger. If the goods have been exposed to prolonged temperatures above certain thresholds by the final destination, the shipment will be rejected.
Solution
The models were based on time series and auxiliary data from hundreds of thousands of historical missions. The approach can be used to optimize the choice of packaging solution and transport route, and act as the basis for a warning system that notifies operators of increased risk along the way. The final solution was deployed in a web app for showcasing and internal use.
Tools
Bayesian regressions were used to estimate distributions in shipment maximum and mean kinetic temperatures. Tree ensemble models built classifiers for probability point estimates, and deep neural networks were used for ahead-of-time predictions of the time series. Preprocessing, modeling, and validation flows were developed in Python with data fetched from a relational database. Deep learning architectures were developed in PyTorch and TensorFlow.
Value created
The models predict the risk that a temperature-sensitive shipment will fail before it does, given a specific transport route and packaging, using time series and auxiliary data from hundreds of thousands of past missions. That lets the client optimize packaging and transport routes up front, and underpins a warning system that can alert operators to rising risk while a shipment is still in transit, rather than discovering a rejected shipment only at the destination. The solution was deployed in a web app for internal use and showcasing.
Learn more
