Predicting traffic light behavior to smooth out traffic in a smart city

Can a cost-efficient cloud-based connection between drivers and traffic lights create smoother traffic and lower emissions? We launched a collaborative pilot program to find out

Our resources for data are growing every day, especially data related to the places where we live. Cities across the world are busily creating open data portals, providing municipal infrastructure information to the public for the purposes of creating better services in urban areas.

When considering what services could be made better using data, we considered that how we navigate through traffic lights was a prime candidate for improvement. To learn what could be accomplished, we launched a collaborative pilot program which included participation from City of Hamburg, SWARCO, and Audi.

Our purpose was two-fold. First, we examined if traffic flow and vehicle emissions were improved when providing drivers with information about traffic lights on the road ahead. Secondly, we learned if the relevant real-time infrastructure data of an 70 intersections could be managed efficiently by a cloud platform, and be delivered to drivers in a timely manner through existing mobile 3G/4G/LTE channels.

In brief: Yes, and Yes. 

But, how did we do this without rewiring every traffic light in Hamburg? That’s where the collaboration comes in, as we also discovered was that no individual company could have accomplished this alone.

To begin, LSA Hamburg provided HERE with all the raw traffic signal data, along with the topology and location of the city’s traffic lights. At HERE, we created highly accurate digital models of some of the most complex intersections in the city of Hamburg. We added those intersection models to the HERE Open Location Platform, along with the raw traffic data – creating a centralized, low-latency cloud-based repository of information.

BLOGHere_hamburgtraffic_body1

By connecting to the HERE Open Location Platform, SWARCO was able to build predictive models based on the combination of the raw data from the city of Hamburg and the HERE intersection models. Those predictions were then provided back into the HERE Open Location Platform cloud and made available to applications in near-real time. Audi provided the vehicles, and the driver applications  

…and it worked.

When provided the condition of oncoming traffic signals, drivers experienced smoother traffic, and their vehicles produced fewer emissions.  Fundamentally, the system reduced unnecessary acceleration, which also reduced braking, which thus creates a smoother flow of traffic. This also optimized the start-stop engine feature (idle engine time) when waiting for a red light to change.

The pilot was also a technological success. Hamburg’s tight network of complex intersections was mapped to to lane-level accuracy standards, and made readily available through a cloud network in real time. The full pilot was conducted without the need to modify any traffic signals in Hamburg, or building expensive temporary infrastructure.  All it took was a little collaboration.

The European-wide C-ITS program seeks to utilize information technology to increase safety and reduce emissions on the roadways. One of the “Day 1” applications that the program seeks is better utilization of city data to inform drivers on the condition of traffic lights, and our collaborative pilot program has proven that it can be provided in a high quality, high-performance experience through complex intersections via a city’s existing mobile communication networks.

Though this program was test, its success proves that these projects are worth more exploration. We’ve seen that emissions and traffic flow can be improved by sending traffic light information to drivers - and we know for a fact that better-informed drivers make for safer and more efficient roadways, as well as a healthier environment

If you could create a test for improving traffic and emissions in your city, what would you design?

Topics: Editor's picks, Smart cities, Open Location Platform, Connected cars

Comments