Tech Talk: Audi, Traffic Light Information And The Future Of What—And How—To Drive

V2I. C-V2X. DSRC. What's it all mean? Why is it more important to you than you might think?

Audi has been on the forefront of vehicle-to-infrastructure (V2I) technologies for more than a decade—all with the aim of creating intelligent, more engaged vehicles that add value to drivers.

Today's technologies include Traffic Light Information (TLI), which is offered with a subscription to Audi connect® PRIME in most Audi vehicles on sale today, and provides a countdown in the instruments of how long a traffic light will remain red before turning green (Time to Green—TTG). TLI can also offer recommended driving speeds (Green Light Optimized Speed Advisory—GLOSA) to limit the number of red lights hit—which also carries the benefit of helping conserve fuel. Connected to infrastructure as well, using TLI, cities will soon be able to aggregate data to see how long cars are waiting at lights and optimize signal timing for better traffic flow.

This interactive map shows where all Traffic Light Information intersections in the U.S. and Canada are located. Initial TLI deployment is often focused in larger Audi sales markets.

Tomorrow's technologies are designed to enhance TLI with more information and faster signals, adopting 5G-powered cellular vehicle-to-everything (C-V2X) technologies that will be able to send messages near instantaneously to and from other vehicles, school zones and worksites. This future technology has the potential to increase driver confidence on the road by providing warnings—or even having the car take autonomous action—when it senses an impending collision. Audi announced in early 2020 that it has begun collaborating with the Virginia Department of Transportation, communications chipmaker Qualcomm and Virginia Tech on a pilot program around construction sites.

Travolution: The beginning

Ever technology-forward, back in the mid-2000s, Audi engineers were looking at how to manage traffic flow and create a communications channel for which they could inform automated driving. Using dedicated short-range communication (DSRC), a direct wireless channel, the engineers worked with the city of Ingolstadt, Germany, Audi's hometown, to set up six traffic signals to relay information to and from an initial two vehicles in a pilot program.

From August 2006 to October 2010, the project grew to 46 traffic signals in large part because of the switch from DRSC to cellular communication via a central server and became part of a wider program called Travolution. Through the study's findings, Audi and Ingolstadt found their prototype for Traffic Light Information helped reduce the time spent at traffic lights by 21% and decreased fuel consumption 17%. By comparison, engineers often spend tens of millions of dollars on lightweighting and engine technologies designed to improve efficiency.

Travolution would expand from Ingolstadt to Verona, Italy; Berlin, Germany; Garmisch-Partenkirchen, Germany; Portland, Oregon; Shanghai, China; and beyond.

After finding benefits with TLI, the next challenges would come with how to grow it on a much larger scale.

Coming to America

With proven results, the project manager for Traffic Light Information at AUDI AG, Michael Zweck, branched out to find a region where it could make sense. Europe and Asia both operated many different types of traffic signals; those across North America were much more standardized, allowing users to benefit in a greater number of cities.

Zweck then met with Thomas Bauer, a longtime transportation and civil engineer in the U.S. and Germany, who had been working with connected infrastructure and automakers. Bauer created a new company, Traffic Technology Services (TTS), to provide value for cities and automakers who could use the information. Together, they officially introduced the world to TLI at the 2014 Consumer Electronics Show in Las Vegas.

Prior to CES, 'The very first proof of concept between Audi and TTS in the U.S. was three signals here in the Portland, Oregon, region,' said Bauer.

'We installed a laptop with two network cards inside their traffic control center,' Continued Bauer. 'It was plugged into their network because it was not connected to the internet. It was a completely insulated network just dealing with the traffic signals. We basically built a fully functional TLI for those signals. Those results are what got Michael Zweck and his colleagues to say, 'Okay, that sounds very promising and doable. Let's see how we can grow this.''

The same year, the team also built a running model in Belmont, California, a Silicon Valley R&D hub for Audi. Greater deployment then began across the U.S.

Predicting traffic lights

As of today, TTS collects data at more than 20,500 intersections within 96 agencies and cities and 26 major metropolitan areas in the U.S. alone. The data—how, when and why lights change color in their grids—is routed back to TTS's servers in Portland. TTS create predictions, running calculations of how traffic lights behave in various scenarios, such as when pedestrians push crosswalk buttons or traffic flow at various points during the day. Using data feeds, engineers are able to adjust for and calculate latency out of predictions.

Once engineers are able to ensure algorithms are within 95% accuracy of the data, even accounting for various traffic situations, Audi and TTS turn on traffic signals for customers to see for TLI, TTG and GLOSA in their cars.

Much the same as with the Travolution pilot, Audi aims to improve convenience for drivers, increase traffic awareness and encourage driving behaviors that promote efficient driving. To do this, Audi and TLI have to precisely predict how traffic lights will behave in the next two minutes in spite of the variability of traffic volume and changing intervals between red and green.

This led TTS to develop a complex analytical algorithm that calculates predictions from three sources: the control program of the traffic signals; real-time data of the traffic computer, a combination of road-occupation cameras, detector strips in the road surface, data on approaching buses and trams and buttons that pedestrians press; and historical data. The forecast algorithm improves itself continuously and learns how, for example, the traffic volume changes in morning commuter traffic or at midday when children leave schools.

Audi vehicles send anonymized data when traffic lights are crossed to a backend, which checks whether the actual crossings of traffic lights correspond to the forecast data. After initial and ongoing quality thresholds are met, the traffic lights are cleared for the display in the car.

The benefit to municipalities for sharing their data is helping improve traffic flow during all hours of the day, future-proofing connected infrastructure and even helping reshape roads based on traffic patterns. TLI can even measure for 'split failures'—when you're at a traffic light and do not make it through in one cycle. Kiel Ova, who is responsible for connecting TTS with municipalities, said data provided back to some cities has helped them alter traffic signal timings to improve traffic flow and create more pedestrian-friendly intersections, which can help cities' walkability scores.

Ova has said that one municipality has even replaced its entire network of 20-year-old traffic signal hardware to jump on board with TLI.

'The mayor wanted to know specifically, from our perspective, what did they need,' Ova said. 'Connectivity is the biggest thing, updating the hardware and software components of the traffic system. Then having human resources, bringing in other departments within the agency to handle requests on a 24/7 basis. It's really neat that we helped a city join the 21st century in terms of connected vehicle technology.'

Where do we go now?

4G LTE signals are able to relay information across the world in a matter of seconds, providing nearly instantaneous appraisals of road traffic, signals and even if and when people press a crosswalk button. What happens when we get to 5G? What happens as more vehicles adopt V2I or C-V2X technologies?

Audi's Michael Zweck believes it will complement the road to future automated driving.

'If you think about automated driving, you usually have two different sensors for the same kind of information. So if you look for road markings, you're using Lidar and you have a camera, and you always verify between these two actions. For traffic lights, you could put their camera into the car,' Zweck said. 'But what is the second physically independent information you could get from a traffic light? That's critical.'

Zweck imagines a future scenario where a vehicle may sense other cars and traffic lights indicating oncoming traffic with an inattentive driver, where the vehicle may warn him or her or even brake if it senses imminent danger. Beyond the potential for fewer accidents, this level of automation and connectivity also carries the potential to lower congestion with reducing particularly inefficient stop-and-go driving, easing traffic flow, creating pedestrian detection at bus stops and, as proven with Travolution, benefitting overall efficiency.

With swarm intelligence—more vehicles on the road—the technology becomes better. With every advancement, communication between vehicles, infrastructure and roadside crews of people helps a vehicle paint a more vivid picture for drivers. Today's technology is designed to help provide drivers with a more detailed look of the road and peace of mind to drivers and to municipalities with data reports that help optimize traffic flow. It also lays the groundwork for tomorrow.

'It is worth the effort to front-loading for that kind of technology,' said Zweck about the decision to put TLI into production. 'You have to prepare because, while it's still in the future, you definitely need to have these in your car and prepare for where we are today and where we are going.'

Photo credit: Audi
posted on conceptcarz.com

Related Posts