Global
Road Infrastructure Manager: Travel Time Service
Road Infrastructure Manager  Travel Time Service    Case Study Mobility
Overview

Development of a data-driven ML model for real-time travel time prediction, based on traffic, flow, speed and density data

Background

Outdated data-driven travel model, lacking transportation logic and sensor network, to be adapted to traffic dynamics. 

Challenge

Training a data-driven ML model based on transportation logic, improving the ability to accurately represent traffic. 

Solution

Approach:

Advanced Machine Learning models with transportation principles lead to develop a more accurate and reliable traffic forecast. Targeted analysis of the most suitable AI architectures and transportation study to correlate flow, speed and density data defined the resulting model. The outcome, based on data-driven techniques, was trained to interpret traffic dynamics and predict the propagation of flows along the network, improving time accuracy and operational efficiency. 

Project Details

Team Involved:
Project Manager (PM) | Subject Matter Expert (SME) – Transportation | Subject Matter Expert (SME) – Machine Learning | Subject Matter Expert (SME) – Google Cloud Technologies | Data Scientist (2x) | Consultant
1/5 phases
Assessment:
Data analysis and related DBs for the routes under examination and considered time horizon. Infrastructure structure: Evaluation of the relationships between data and client systems.
2/5 phases
Data preparation
In-depth data study, carrying out data cleaning operations, outliers identification and pre-processing of the available data.
3/5 phases
Model development
Model training defining the main features to be taken into account
4/5 phases
Model evaluation:
Evaluation of the quality and ability of the model to represent reality through statistical time-series.
5/5 phases
Distribution and monitoring:
Providing the model in a specific Google environment and maintenance over time.
Insights and Thoughts

Best Practices:

"The integration of IoT and analytics proved essential for optimizing production processes."

Data-driven decisions increased by

+30%

ROI achieved within the first year.

150%

Carbon footprint
reduced by

+25%

Waste production
decreased by

-22%

Ready to collaborate and create value?

Let’s get in touch and see a new vision of the future!

Enter your details, one of our experts will get in touch soon