Better weather monitoring could prevent road crashes in fog

A system that can detect the fog and predict it would be beneficial in the UAE.

ABU DHABI // A system to monitor and predict adverse weather conditions in real time could help prevent crashes.

“For the past few days, we have had dense fog here,” Dr Mohammed Abdel-Aty, chairman of civil engineering and construction engineering department at the University of Central Florida, said on the sidelines of the International Road Traffic Accident Conference in Abu Dhabi.

“A visibility sensor is not enough, but what is needed is a system that will not only enable us to detect the fog but for us to expect when fog will happen. We have to look at transport and traffic management proactively with real-time prediction.”

Dr Abdel-Aty is working on a project using data safety analytics.

“Fog is coming from different parameters but what we are more interested in is not the result of reduced visibility but what caused the visibility problem,” Dr Abdel-Aty said.“These include temperature, wind speed, the moisture on and above the ground, the elevation and topography so if you have a dip, it is likely that fog will settle. If we are able to detect these parameters, we can predict the fog and the level of fog.

“For the first time, we are looking how we can predict fog on the roads at least 15 minutes in advance to take the necessary measures. If fog is dense, we can close the road, and if it isn’t that dense, we can put advisory speeds to reduce the speed.”

The benefit of this system is to prepare emergency and traffic management.

“If fog descends suddenly, you just need one driver who is speeding under reduced visibility and once accident to trigger a multiple-car pile-up,” Dr Abdel-Aty said. “We want to be a little early before fog occurs.”

The approach is to look into the weather, how the change of visibility affects driver behaviour, and the distance between vehicles.

“In our project, we started to see that even without the visibility detectors, if we just monitor the speed and the space between you and the car in front, you can tell that visibility is starting to deteriorate,” he said.

“We can continuously collect data on the road – the number of vehicles, the spacing of vehicles, and vehicle type every few seconds. And if you put these with the road characteristic with the weather and land use data, it becomes what we call ‘big data’.”

This “big data” can be transferred to a traffic management centre, which processes the data and evaluates the risks.

“We have real-time data and we encourage countries to use it,” he said. “We have more roads, more drivers and more cars so we need to develop systems to predict the problem.”