| CPC G01C 21/3691 (2013.01) [G01C 21/30 (2013.01); G06F 16/285 (2019.01); G06F 16/29 (2019.01); G08G 1/0112 (2013.01); G08G 1/0129 (2013.01); G08G 1/052 (2013.01)] | 15 Claims |

|
1. A method for detecting and obtaining lane-level insight in unplanned incidents, the method comprising:
acquiring, from a plurality of probe vehicles, probe data for a road strand comprising a plurality of lanes;
map matching, using a lane level map matcher, the probe data;
splitting the road strand into a plurality of lane strands;
generating, using a clustering algorithm and the map matched probe data, three or more clusters of vehicles each comprising one or more probe vehicles on a same lane strand of the plurality of lane strands within a threshold distance of other probe vehicles in a respective cluster, wherein at least two of the three or more clusters of vehicles are located on different lane strands of the plurality of lane strands;
determining an average speed for each cluster of the three or more clusters;
sorting each of the three of more clusters in a sequential order using an x-value in a center of each of the three or more cluster;
calculating a speed-ratio for each contiguous pair of clusters of the three or more clusters where the speed-ratio=(average speed of a first cluster and a next sequential cluster)/(average speed of the first cluster);
computing a JAM factor for each of the three or more clusters, wherein the JAM factor comprises a speed-ratio of the average speed vs a free flow speed of the road strand;
calculating a LOS delta value or a change in a JAM factor across K clusters of the three or more clusters using LOS delta value=JAM.Factor of k+1−JAM.Factor of k, wherein if the LOS delta value is greater than a predefined threshold, then a cluster k is determined to have an unplanned traffic incident.
|