Real-time Safety Alerts for Severe Weather and Jam Tails



Real-time Safety Alerts for Severe Weather and Jam Tails

Authors

Maarten Clements, TomTom, Nick Cohn, TomTom

Description

We present a system to derive the impact of weather on road congestion, using TomTom floating car data. This enables the creation of alert messages on heavy weather and prediction of the location of jam tails for user devices or road authorities.

Abstract

1. Challenges and objectives
Unexpected traffic jam tails and severe weather conditions are among the most common causes of road accidents. The majority of drivers have experienced situations where a sudden change in speed or weather conditions has resulted in dangerous situations. Our challenge is to automatically provide reliable safety alert messages derived from floating car data and real-time weather polygons. It is absolutely critical to deliver these messages with great accuracy to the right driver at the right time.

2. State of the art (description of the situation, existing data)
Floating Car Data has become an increasingly important and reliable source to derive congestion information. However, accurate delivery of alert messages at the right time and location remains a challenge. The availability of highly accurate floating car data and weather data now allows for detailed analysis on these dangerous road situations.

3. Methodologies, ideas, techniques and innovative methods
By analyzing an extensive collection of floating car data and correlating this data to real-time weather information we have derived a speed adaptation model. The model predicts the impact of severe weather conditions on the actual driving speed on the road. The model can thus be used to combine weather and traffic information to provide accurate alerts to the driver. On top of this, the detail of the information allows for accurate prediction of the position of the jam tail, the location where cars are most likely to collide.

4. Theoretical or experimental results and interpretations
Our analysis shows that in many cases we can accurately predict the tail of a traffic jam and if needed provide the current weather conditions as additional information to the driver. This has been evaluated by comparing the provided alerts to the actual road situation.

5. Deployment and ways forward
The derived weather and jam tail alerts are delivered to car drivers on their navigation device. Whenever the driver approaches a traffic jam with high speed, the device alerts the driver to slow down. In areas with bad weather and heavy traffic, the driver receives information on traffic jams that are likely to be impacted by the weather. Besides the common delay information the alert will contain information on the type of weather (snow, rain, …).
Next to direct delivery to the navigation device, a service has been set up to deliver the alert messages as direct data stream to road authorities or governments, who can alert drivers over common communication channels like websites, radio and dynamic traffic signs on the road.
Future development will focus on providing predictive information using the weather forecast and further fine-tuning of the alert accuracy.

6. Conclusions
We present a method to derive the impact of the current weather conditions to real-time road congestion. The model allows creating accurate alert messages on weather related congestion and the actual location of jam tails. The alerts are provided to connected navigation devices and road authorities to provide safety alert messages at the right time to the impacted drivers on the road.

Publisher

Association for European Transport