Smarter Infrastructure Is the Key to Safer, Smoother Mobility
February 06, 2023
An internet search of “highway pileup” returns multiple articles of deadly accidents, many of which report reduced visibility due to weather conditions as the leading cause. What if the lack of visibility could be overcome by the highway itself communicating this incident to all cars approaching the accident area?
This future doesn’t have to be science fiction. Smart infrastructure can make mobility better and safer – helping us get to our destinations faster with adaptive traffic lights, hassle-free travel in and out of crowded parking lots after events, and reducing accident risks as nearby collisions are immediately communicated to our vehicle.
In this article, we’ll explore the evolution of Vehicle to Infrastructure (V2I) and how ubiquitous, advanced computing can usher in a better, safer era of transportation.
So, what is V2I?
Vehicle-to-infrastructure (V2I) is a communication model that allows vehicles to share information with the components that support a country's highway system – such as traffic lights, lane markers, streetlights, signage, overhead RFID readers, cameras, and parking meters. Nearly any object, from buildings to light posts to residential neighborhoods, can be equipped with this capability.
It essentially gives vehicles a new level of intelligence which, when combined with smart roadway components, can make driving safer and more predictable.
Examples of the current state of infrastructure-to-vehicle communication include:
- Municipalities using intersection cameras and radar systems to control traffic lights manually.
- Vehicle owners collecting parking tickets upon entry and pay before exiting after spending some time locating their vehicles in big parking lots like those found in airports.
- Municipalities using traffic cameras to notify central command about traffic flow and then in turn update digital signage to reflect slowdowns or stoppages ahead.
- Users utilizing navigation and map applications.
But as the technology matures, we may see game-changing improvements to mobility, including:
- Smart parking lots where the infrastructure can serve as an intelligent operator, handling ticketing and payments, while also helping owners easily locate their vehicles in massive parking garages.
- On-demand traffic signal priority to reduce traffic jams or automatically detect and help clear the way for an ambulance.
- Real-time accident reports or Local Hazard Information (LHI) to notify drivers in advance of dangerous road conditions (floods, ice) or pileups ahead of a curve.
Many automakers and cities have already begun implementing V2I in the real world with great results. Yet there is still much more that can be done to improve capabilities and accelerate global adoption.
Two Different Approaches and Opportunities
Today, there are two primary types of V2I implementation - Dedicated Short-Range Communication (DSRC) and Cellular Vehicle-to-Everything (C-V2X).
DSRC allows nearby vehicles, RSU (Roadside Units), and pedestrians to share information with low latency and high reliability with each other. It is based on Wi-Fi and uses standard IEEE 802.11p, which was approved in 2010, and is primarily focused on safety-critical applications.
This technology is proven and validated today but will require that vehicles be equipped with an On-Board Unit (OBU) and requires that infrastructure be upgraded with RSU installations. That said, once investments have been made, this system can often operate at a lower cost than a pure cellular-based approach.
C-V2X takes a similar approach but involves cellular connectivity as an intermediary. Vehicles will be able to communicate with transportation infrastructure, other vehicles, the cloud, cellular network infrastructure, and pedestrians, with computing handled by a user’s phone vs an OBU.
While many vehicles come with cellular connectivity today, infrastructure investment is still needed to be able to process data from an increased range. Increased adoption of 5G will likely cause the support for C-V2X to grow, but the likely higher cost (due to network usage) is a potential downside.
V2I RSUs Requirements Are Evolving
A typical V2I RSU currently boasts the following features:
- Low latency to deliver critical information on time
- High Reliability to deliver critical information without errors
- Cybersecurity to avoid compromise of road safety and user information
- Control system capability to perform local operations like opening gates
- LTE/5G for network connectivity
- Interface with a wide range of sensors like cameras, LiDARs, radars, etc.
- AI/ML inference capability to optimize traffic, recognize vehicles, etc.
V2I RSU requirements are constantly changing based on the application use case it is being designed to be deployed in. For realizing these requirements, you need a hardware platform that is adaptable to implement today’s needs but also future-ready to take on tomorrow’s needs. AMD Adaptive Computing platforms are heterogenous hardware platforms that are made up of application processors, real-time processors, video encryption and decryption, AI/ML inference (Deep Learning Processing Unit), Security Features, and programmable logic that can be used to implement almost any type of sensor. Adaptive Computing platforms are flexible enough to process the tremendous amounts of data generated by these systems and feed V2I communications with real-time updates.
Advancements in Programmable Logic
It is worth diving into the computing power needed to make smart cities run. Most people are aware of the importance of semiconductors as the COVID-caused chip shortages disrupted several aspects of our daily lives – from electronics to vehicles and more.
Smart cities will also rely on these chips for processing power, but we’re well past the one-size-fits-all approach.
Customized solutions are growing in popularity as they can be tailored to each task. Field Programmable Gate Arrays (FPGAs) are well-suited for the changing needs of smart cities because, unlike pre-made chips, they can be reconfigured multiple times. Engineers can adjust the I/O path and memory to suit the needs of the smart city device, be it a traffic light, accident detection sensor, or parking structure – helping make each device faster and more energy efficient.
This adaptability also allows for the chips, which are the brain in each device, to be reused which can save the city money versus pre-programmed solutions which only serve one dedicated purpose.
Will We Enter a New Era of Mobility?
With the many advancements in tech-enabled transportation, it shouldn’t be a surprise that governments worldwide are investing in smart infrastructure. Intelligent mobility will be an important piece in enabling these smart cities due to the increase in safety, reduction in traffic, and connectivity it can provide all citizens.
Achieving the full potential won’t happen without concerted effort and investment. Governments should embed RSUs and other sensors seamlessly into the roadway infrastructure, which is already being seen in FasTrak lanes in the Bay Area or smart traffic lights for EMTs and firefighters. But we’re still in the early stages of the journey.
These smart devices can provide the ability to monitor traffic patterns like never before. For the first time, we can have a 30,000-foot view of cities in their entirety, allowing us to identify and fix dangerous intersections and revamp older, existing infrastructure and public transportation in ways that better serve current residents.
I believe that transportation in the next decade will be almost unrecognizable from a technological perspective. Advancements in sensing, computing, and short-range communication can make driving safer and better for the next generation.