IceWay

Logo
Roadmap of Iceland with cameras
Snow recognition AI

Inspiration

The hostile and unpredictable weather conditions in Iceland present a considerable challenge to road maintenance and safety. Our team was inspired by the potential of harnessing diverse datasets and AI to provide more accurate, real-time insights into road conditions. The vision of "IceWay" emerged from a desire to significantly reduce the man-hours required to monitor road conditions while optimizing resource allocation for snow clearing, which is not only time-consuming but also financially draining. By accurately predicting road closures or clearances needed, we aim to contribute to better road safety, reduced costs, and improved efficiency in dealing with snow removal and road maintenance.

What it does

IceWay integrates datasets from various sources including weather data from meteorological agencies, road condition information from the Road and Coastal Administration, and city data. By employing AI, it analyzes this data to better predict road conditions. Utilizing historical weather data and road conditions, an AI model is trained to extrapolate road conditions given certain weather parameters. The model then uses weather forecasts to predict road closures or clearances needed, allowing for optimized allocation of snow clearing resources. Furthermore, real-time data is provided to the public and authorities, enabling better planning and decision-making.

How we built it

The building process of IceWay involves several methodical steps:

Representing the Icelandic road network as a computational graph using the networkX python module.
Collecting weather conditions, traffic data, and road camera images from various sources like vedur.is, openweathermap.org/api, and gagnaveita.vegagerdin.is/api.
Training an AI model on historical weather and road conditions data to extrapolate road conditions given weather forecasts.
Developing an algorithm for optimally allocating snow clearing resources. This includes employing a Convolutional U-net architecture for semantic segmentation to detect snow from camera images, and utilizing a Naive Bayes classifier for predicting road closures/clearances by considering the joint probability of snow-fall and traffic.

Challenges we ran into

One of the primary challenges was the accurate integration and interpretation of varied data sources. Ensuring the robustness and accuracy of the AI model in the face of Iceland's volatile weather conditions was another hurdle. Additionally, devising an algorithm that could efficiently allocate snow clearing resources while considering multiple variables proved to be a complex task.

Accomplishments that we're proud of

We are proud of developing a solution that could potentially have a significant positive impact on the community by improving road safety and reducing operational costs. The successful integration of diverse datasets and the creation of a robust AI model to provide accurate road condition predictions are major accomplishments. Furthermore, the project's potential to contribute to environmental sustainability by reducing fuel consumption and greenhouse gas emissions is an achievement we value greatly.

What we learned

We gained valuable insights into the intricacies of data integration, machine learning, and real-time prediction. The process of collecting, cleaning, and utilizing data from various sources was a significant learning experience. Additionally, we honed our skills in applying machine learning algorithms to interpret the complex relationship between weather conditions and road states.

What's next for IceWay

Looking ahead, we envision further refining the AI model to improve its accuracy and expanding the data integration to include more datasets for a comprehensive analysis. We also plan to explore partnerships with local authorities and other stakeholders to implement IceWay as a practical solution for road maintenance and safety. By continuously improving and adapting the system, we aim to make a lasting contribution to road safety, environmental sustainability, and the overall efficiency of road maintenance operations in Iceland.