Forecasting Overview
Advanced forecasting solution transforming fuel inventory management through predictive algorithms, comprehensive tracking, and real-time optimization for downstream oil and gas operations.
Written by Jess Hamilton
Updated at October 10th, 2025
Table of Contents
How does our forecasting engine work?
1. Configuration: Settings specified by the client determine how data is processed and presented.
2. Inputs: Inventory and consumption data are captured automatically or provided manually
3. Real-Time Analysis: Continuous monitoring of inventory and consumption
4. Results: Actionable insights are provided throughout the platform
The forecasting engine in Titan Logistics (TFX) leverages inventory data and delivery details to determine when and how much fuel a site will need.
How does our forecasting engine work?
1. Configuration: Settings specified by the client determine how data is processed and presented.
- Type of forecasting enabled:
- Inventory Threshold Based Forecasting: Reactive. Orders are generated based on the level of product in the tank.
- Rate of Consumption Based Forecasting: Proactive. Orders are generated based on historical product consumption rate.
- Level of forecasting enabled:
- Product Level
- Tank Level
- Prediction window (1-7 days)
- Product type(s) or designated product mix
- Delivery buffer windows
- Trailer proxies: Compartment / Capacity
- Mass Tank Updater
- (Optional) Manual entry of Consumption Pattern
2. Inputs: Inventory and consumption data are captured automatically or provided manually
- ATG Integration (15+ options)
- API-based inventory readings
- Manual inventory uploads
- Event, holiday, and weather considerations
- AI-driven consumption pattern recognition
3. Real-Time Analysis: Continuous monitoring of inventory and consumption
- Analysis triggered by every polling or configuration change
- Frequent polling intervals (10-30 minutes) is supported
- Analysis of short-term (4 weeks) and long-term (previous seasons) historical data
- Product-specific logic and prioritization
- Primary Product: The prioritized product type
- Secondary Product: Another product that may be used to complete a full truckload
4. Results: Actionable insights are provided throughout the platform
- Predictions displayed by order or by full truck load
- Priorities are automatically assigned
- Must Go: Runout will occur if delivery is not made in the CURRENT shift
- Should Go: Runout will occur if delivery is not made in the NEXT shift
- Could Go: Compartment/FTL can fit without a retain
- Delivery window elements are presented consistently:
- Retain: Too early
- ETA: Ideal Time
- Runout: Too late
- Predictions are modifiable
- Results may be sent directly to dispatch or to a unique end point