AustraliaThe new project in Melbourne will use AI, deep learning algorithms and predictive modeling to reduce travel times, emissions and traffic congestion.
“Smart Corridor” is the world’s first traffic optimization system pilot project. Photo: Kapsch
The “Smart Corridor” is a three-year trial on a 2.5km stretch of road in Nicholson Street, Carlton, one of Melbourne’s busiest streets. The University of Melbourne partnered with Austrian company Kapsch TrafficCom and Transport Victoria to carry out the project. The traffic management system launched at the end of March uses historical and real-time data from a huge and diverse network of sensors, including CCTV cameras, Bluetooth, quality tracking devices air traffic, transit information, TomTom service data, weather data, traffic light signals, intersection logic data.
Some data is available in the city, others are pre-installed in the Australian Integrated Multimodal Ecosystem (AIMES) project. AIMES is the world’s first and largest ecosystem for large-scale testing of new linked transport technology in complex urban environments.
The project team says this is the first time historical data and real-time data have been combined in a traffic management project, processed by artificial intelligence (AI) and learning algorithms. deep. First, the system can control traffic lights at every intersection on the corridor to achieve optimal traffic flow. “We use images captured from one of the hundreds of thousands of cameras across the network. We use a flexible deep learning platform to analyze images, then create perceptions. We review data as detects long lines of traffic in a lane. That affects the phases and transitions of the lights. We can adjust and optimize traffic at the intersection,” said David Bolt, vice president Kapsch company president, explained.
The system has many ways to communicate with road users and public transport, to influence traffic flow in the event of an accident or to balance vehicles. If an accident prevents a tram from crossing an intersection, the system can match each oncoming tram to suggest that it carry stranded passengers to their destination.
Traffic management systems can monitor pedestrian crossings and respond to drivers via communication with connected vehicles. One example the team gave was when running through a special intersection where the driver had to steer through a tight bend and couldn’t see the pedestrians until he was about to hit them. Real-time alerts will be sent to connected cars if the system detects the driver is about to hit a bend like this.
Accident management is also an important part of the system. The new system will alert the operator when it detects an anomaly or predicts an impending problem. Operators can choose from a menu of actions to handle the situation, or review real-time camera images to determine what’s happening. They can also look for similar accidents in the area’s history, including effects on traffic.
Over the next three years, the team hopes to test everything from making sure emergency vehicles only see green lights, to redirecting traffic around the school at pick-up/drop-off times, to changing traffic routes. information according to the air quality map, automatically texting the car owner parked in a position causing traffic jams. The project will collect data before and after use to measure and monitor the effectiveness of the system. According to Kapsch, the system is designed to scale from small intersections and short-distance deployments to widespread citywide adoption.
An Khang (According to New Atlas)
at Blogtuan.info – Source: vnexpress.net – Read the original article here
