NTT computor & data science institutes devised “Aggregation sudden outbreak index calculation technology”. This technology have been applied to the traffic congestion detection technology classified by the lane.

The traffic congestion detection technology classified by the lane is a technology which detects the motorcade which occurs per lane. It assumes the spread of the connected car.
For example, it detecst the traffic congestion which occurred only on the specific lane from the image of dashcam. It optimizes traffic by this.

It assumes that the left side of road is congested. A driver does not know to be traffic congestion of the vehicle located whether it is in a line with drive-through (α in the illustration) or in a line toward the home center (β in the illustration).

The driver which wants to go to home center will notice that it is traffic congestion in a line with the drive-through on the way.
When it turned out that it is traffic congestion of drive-through in advance, this driver did not need to be involved in traffic congestion. If the mistake decreases, traffic congestion of drive-through will also be eased.

It collects and analyzes the image of dashcam which is near the proposed site of traffic congestion in the traffic congestion detection technology classified by the lane.
Aggregation sudden outbreak index calculation technology is needed here.

Huge traffic is required in order to upload the image of dashcam to the server and to analyze it. It also requires huge computational complexity because the image analysis is so heavy.Therefore, it is better to process only required image preferentially.

Aggregation sudden outbreak index calculation technology can find a location crowded with few amount of datas, controlling computational complexity.