Case studies/Analytics Case-Study

Forecast-based tour planning.

Energy logistics · 2023

Challenge

Demand at individual filling stations fluctuates strongly, while many operational constraints for vehicles, staff, compartments, and roads must be respected. Costs needed to fall while customer satisfaction stayed maximised.

Solution

A station-specific forecast model, including Prophet, DeepAR, XGBoost, and an LSTM, was combined with a mathematical optimisation algorithm for tour planning, including Google Maps API routing data.

Impact

Over 10% efficiency gain through better planning and the saving of 5 full-time roles through automation.

Ready when you are

Get your time back.

Stop the busywork. We automate the mundane so you can focus on what matters.

Updates on AI, delivery, and Askantis — unsubscribe anytime.