Apple might not want it, but a group of Israeli researchers certainly like crowdsourced navigation app Waze. Using data submitted by Waze users in Israel, the team from Ben-Gurion University uncovered some findings that, in theory at least, could signal to police and public-safety officials where they might best invest resources to prevent further damage.
The researchers studied more than 5,000 accident reports as well as nearly 30,000 reports of police presence to determine where the most accidents happen and whether police were often nearby. What they found was that 75 percent of the most-accident-prone areas were intersections, and that there were many instances where police were monitoring safer intersections and weren’t subsequently reported as being at accident sites shortly after they occurred. When police did respond, the average response time was just over 28 minutes.
Satellite image from Google Earth combined with accident and police report heatmap – area units with high accident scores (marked as redand white ellipses) and high police scores (marked as purple circles). Source: IEEE 27th Convention of Electrical and Electronics Engineers
According to the paper, future directions for this research include analyzing the effect of greater police presence on the number accidents and “us[ing] machine learning techniques to understand the causes that make some intersections more dangerous than others.”
These researchers are just one of many groups trying to make sense of traffic data to improve our commuting lives, though. A few that we’ve covered recently are INRIX, which collects real-time traffic data from millions of cars, commercial vehicles and mobile phones; Xerox, which is trying to quell congestion on the streets and freeways of Los Angeles; and Nokia, which is experimenting with a credit-based system for finding and reporting open parking spots.
Feature image courtesy of Shutterstock user isaravut.
 |
 |