Artificial Intelligence and 5G to Ease Global Traffic Woes With Intelligent Transport Systems

Amitabh Ray, Managing Director, EricssonIf we sent up a drone camera to take pictures of Indian roads during peak traffic it would resemble a mass of unmoving metal with bumper-to-bumper traffic gridlocking every major road. The situation is so bad and turning worse each day that people point out Bangalore’s most infamous logjam at Silk Board Junction which has inspired its own Twitter parody account as “India’s largest parking lot.” Traffic snarl ups not only causes frustrations to millions of passengers, the estimated financial cost is over $2000 billion per year worldwide.

But it’s not Bangalore, Mumbai or New Delhi that takes the crown for the most gridlocked city in the world it’s Los Angles.The Global Traffic Scorecard, from INRIX, a transportation analytics company, found that LA drivers spend an average of 102 peak hours in congestion 11 hours longer than the next most congested cities, New York and Moscow. Sao Paulo’s drivers sit in fourth place, with 86 hours spent in traffic while San Francisco ranks fifth with 83 hours. Out of the 1,360 cities analyzed, London was seventh, behind Bogota who had 75 peak hours of congestion. On average in the UK, drivers spend an average of 31 hours a year in traffic which in 2017led to a loss of more than £37.7 billion in related costs.

The easiest solution we often think is to spend money on infrastructure, but that’s not a silver bullet. The solution lies in a combination of measures from remote working, to spreading the business districts, ride sharing, road user pricing and of course Intelligent Transport Solutions(ITS) such as dynamic traffic lights, the wider use of all lane running on motorways and the efficient planning of road works.

The best starting point to implement an ITS is with collecting data and running algorithms with it to find out the source of the current problem and forecasting what will happen
next. If we know what can be the possible areas of bottlenecks in the next few hours, the traffic management system can be pressed into action to avoid the gridlocks.

Longer range forecasting will help policymakers and transportation professionals to know when and where congestion is worst to prioritize investments with limited budgets. Scientists at Nanyang Technological University in Singapore have developed a new intelligent routing algorithm that attempts to minimize the occurrence of spontaneous traffic jams.

A team from the Texas Advanced Computing Center (TACC), has demonstrated that AI can help to optimize traffic flow. Their work is on developing searchable traffic analyses using deep learning and data mining. The tool uses the raw data generated from traffic cameras, with their algorithm capable of recognizing the various objects in the footage and then characterizing how those objects move and interact. This information can then be queried by traffic planners to better understand the transport network.

The system automatically tags each object it encounters in the raw data and then tracks their movements throughout the footage. By comparing outputs from each frame, the algorithm discovers relationships among the objects. The system was tested in two practical use cases. The first saw it count the number of vehicles traveling down a road while the second identified near misses between vehicles and pedestrians. The results showed that the algorithm was around 95% accurate when counting the vehicles.

According to a McKinsey report Artificial Intelligence and Machine Learning will be the technological foundation for both ITS and autonomous vehicles. The convergence of Internet of Things in connected cars, autonomous vehicles, and sensor based traffic management systems will become a reality with the launch of 5G. One of the key advantages of 5G is low latency which will help autonomous vehicles to respond in real time to dynamic situations.

The lower latency and high throughput of 5G will support connected cars, transportation and retail logistics that consist of fleets ofconnected/driverless vehicles transporting people and goods. The key network requirements for missioncritical automotive driving are high throughput and low latency up to 100 milliseconds. A journey from A to B in a driverless vehicle could involve vehicle-to-vehicle connections, connections between vehicles and street infrastructure for traffic management, and high-speed reliable connectivity to support cloud applications which will be made possible by 5G. Failure is not an option in these cases.

These technological advancements are making possible improved transport. The governments across the globe are also taking initiatives to focus on research of cutting edge technology related to advanced products such as vehicle adhoc networks. These major trends are poised to fuel the growth of the global market for intelligent transport systems which according to estimates is likely to cross $30 billion by 2022.