)
)
)
How can we help you?
)
Marija Mijic
HR & Office Allrounder
T +43 660 383 95 46
office@craftworks.at
craftworks GmbH
Schottenfeldgasse 20/6a
1070 Vienna
Optimized ETA and Routing
with Predictive Analytics
Optimize ETA precision with AI solutions in manufacturing, ensuring timely deliveries and streamlined operations.
:quality(80))
Success Story
reading time: 3 min
Automotive
Mechanical Engineering
Predictive Quality
Industrial Software
98%
Accuracy in ETA prediction
2
Separate predictions on routing and ETA
Freight shipments
Optimized for maximum efficiency
Context
Rail Cargo is the freight shipment arm of the Austrian Federal railways. Freight shipments can often be delayed and complex route routing makes it difficult to accurately estimate when shipments would arrive.
Challenge
The data from shipments had to be evaluated to accurately predict when shipments could arrive at their destination.
Assignment
craftworks had to train a model based upon relevant data in order to accurately predict the Estimated Time of Arrival (ETA).
Solution
craftworks used machine learning technologies in order to accurately predict the arrival times of good shipments along railways. The technologies used here were as follows: Backend (Maven, Java, PostgreSQL), Frontend (Angular as Framework, REST as API and Docker as Container) and Machine Learning (TensorFlow, Python as programming language and Docker as container).
The data was collected, compiled, and visualized via a homogenized pipeline, to be used as a basis for training and evaluating a deep learning model. The model was able to make two separate predictions: one regarding the Route and one regarding the estimated time of arrival (ETA).
With craftworks the collaboration is not only exciting and entertaining for us, but also characterized by complete openness. This allows us to form one team consisting of experts from our business units and craftworks experts. We got surprised that after only 5 days, a fully functional prototype was available. Our employees of RailCargo want to continue working on it and further develop a pilot with craftworks - we are very much looking forward to it!
Peter Schindlecker
ÖBB, Innovation Manager
:quality(80))