All the latest quantum computer articles

See the latest stories on quantum computing from eeNews Europe

Saturday, April 07, 2018

Live streaming traffic info uses Intelligent Edge Analytics

By Nick Flaherty

A live streaming traffic information service in the US is being powered by machine-learning and edge computing.

Trafficware has teamed up with Silicon Valley edge intelligence software firm for the TidalWave service. The service is based on work in Palo Alto in 2015 to provide the first open source traffic data used in connected vehicle applications. This provides streaming traffic data with sub-second accuracy using edge computing combined with a more affordable cloud service with low overhead and so there is no impact to city infrastructure.

Today, the majority of routing and logistics applications rely on historical cellular GPS data to measure roadway congestion and estimate travel times. In order to determine traffic congestion on arterial corridors, the applications assume that all cell phones are located in moving vehicles and reflect current conditions. The speed and accuracy at which the data is collected, analyzed and made available is slow and often does not reflect the actual experience of drivers.

TidalWave performs the traffic and signal analysis either at a city’s advanced traffic management system or on controllers at street level and generates highly accurate real-time information. The efficiency of the edge solution means that data volumes are reduced by a factor of over 100 and can provide hardware savings of up to 80% compared to traditional solutions. The service is a simple software addition to existing city infrastructure and subscribers to the Tidalwave service receive traffic information from a real-time API.

“TidalWave was designed using the first of its kind, ‘intelligent edge’ architecture for use by the connected vehicle, smart cities, and ‘Internet of Things’ (IoT) markets, and will lead Trafficware and the ITS market into the next transformative era of technology over the next decade,” said Joe Custer, CFO of Trafficware,

"TidalWave analyzes, learns and predicts as data is created, at the edge, on existing hardware using a powerful edge compute/data fabric. It delivers precise, granular traffic data at a resolution of hundreds of milliseconds, at a small fraction of the cost of central cloud-hosted learning and prediction,” said Rusty Cumpston, CEO of SWIM.

No comments: