The iconic movie “The Fifth Element” showed us many futuristic innovations. Probably the most bizarre and fascinating was three-dimensional traffic in vertical megacities. Will we ever get to that state?
Perhaps. But if we are ever to make the leap, we must first figure out how to effectively coordinate traffic in 2D. Given the rapid growth of transport modes and urban populations, that’s already an admirable task.
Traffic management systems today are under pressure to evolve and become leaner, greener, and more connected. But what type of technology goes into designing such systems? This post provides details.
Elements of an intelligent traffic management system
Once upon a time, the swing of a traffic safety baton was enough to manage a busy intersection. But with more vehicles on the road, we need smarter systems.
An advanced traffic management system (TMS) is a context-aware solution that relies on real-time data from connected road infrastructure and predictive analytics to effectively coordinate traffic across city arteries.
Such traffic management software, coupled with wireless urban connectivity, acts as a backbone for the implementation of an intelligent transportation management system.
An intelligent transportation system (ITS) is a collection of operational controls and user-facing solutions for coordinating the effective movement of people and goods across different modalities.
ITS traffic systems focus specifically on improving the throughput and safety of urban roads through adaptive controls and analytics.
But as the last 50+ years of targeted effort have proven, managing traffic congestion isn’t an easy task. From city layouts to unplanned protests, a lot of familiar and unpredictable factors make urban transportation planning an uphill battle.
As a concept, intelligent traffic systems were designed to provide traffic managers with real-time and predictive insights about traffic flow speeds and traffic congestion/incidents. In practice, however, the success of such projects strongly depends on a city’s ability to place a virtual management layer on top of physical traffic infrastructure.
Source: IEEE — Smart Town Traffic Management System Using LoRa and Machine Learning Mechanism
Main elements of an intelligent traffic system
|
Hardware
|
Software
|
|---|---|
IoT road sensors including:
Air quality sensors |
Cloud computing and edge processing capabilities:
All supporting transportation apps |
| Connected CCTV cameras |
Big data and predictive analytics
|
|
Connected traffic light systems
|
AI/ML:
|
|
Smart toll gates / electronic road pricing gantry systems
|
Location-based services
|
|
Edge devices — chips on edge nodes for faster data processing
|
Implementing all of the above is a lot of work.
But this progressive digitization also opens opportunities for startups and private companies to pitch their intelligent road traffic solutions to authorities.
6 features global cities seek in traffic management systems
The need for transportation won’t abate any time soon (unless we rewrite the fundamental laws of physics and invent teleportation). But even in that unlikely scenario, someone will have to transport and install the portals around town.
More realistically, the latest prognosis on population growth suggests that urban areas will become even denser and span into megacities with population counts of 20+ million in the next 50 years.
And all those people will want to have multimodal transportation options. McKinsey estimates that between now and 2040, approximately $2 trillion in transport infrastructure investments will be needed every year.
These investments will go into:
- Expansion and modernization of physical road infrastructure — road connectivity projects, electric charging infrastructure development, and maintenance
- Public transportation sector decarbonization and improvement, paired with urban mobility as a service solutions
- Logistics and freight route improvements and better solutions for the booming last-mile delivery sector
- Smart city traffic management solutions to battle congestion and pollution on the streets
Here are the top six intelligent traffic management capabilities urban planners will seek to acquire in the next decades.
1. Video traffic detection systems with edge processing capabilities
Traffic management is prone to the butterfly effect — a phenomenon in chaotic systems (urban traffic) where a tiny change in conditions (one jaywalker) can cause a ripple effect across the entire system (major congestion across town).
Urban planners need more eyes around town to:
- Get real-time information on traffic conditions
- Detect and rapidly respond to incidents
- Proactively implement preventive measures
You can’t catch all the butterflies (or force everyone to respect the traffic rules). But you can create an environment where one casual incident doesn’t bring the entire traffic system to a standstill.
One viable way to accomplish the above is to implement connected video detection systems in critical areas around town. Then pair them with real-time traffic management systems.
Modern traffic incident management systems (TIMs) are powered by:
- Connected CCTV cameras with HD footage
- Computer vision capabilities for image detection and recognition
- Edge chips for local video processing, which reduces latency
- Cloud connectivity + GPS-based communication to receive updates
Such a setup allows you to:
- Detect incidents when they happen — car crashes, road blockages, illegal parking, careless bike riders or pedestrians
- Transmit alerts to the intelligent traffic management system in seconds
- Program or automatically execute a sequence of follow-up actions — dispatch emergency services, adjust traffic signal controls in the area, re-route public transport, and update nearby drivers
The Canadian startup Miovision tested a similar system in the Waterloo region near Ontario. The solution provided local traffic managers with real-time traffic information in the monitored area. Minutes after detecting a collision in a busy intersection, the system alerted urban planners and provided them with real-time footage from the event. The team could then issue a rapid response and reroute traffic from the affected area to prevent further congestion.
At Intellias, we also worked on an IoT-enabled traffic management solution for one of our clients. The pilot system we designed aggregates data from in-car sensors, road cameras, public traffic feeds, and user devices. After being processed locally on the edge device, this information is dispatched to the cloud system for further analysis. Then it is made available to road users and regional traffic management centers as real-time updates on traffic conditions.
Arguably, the best part about building out such edge data processing capabilities (paired with live video) is that you can reuse the collected data for other intelligent traffic analytics use cases. These include:
- Multimodal traffic counts to understand the most-used modalities in the area and their average cruising speeds
- Road safety analytics — using pattern detection, your systems can flag inappropriate driver and pedestrian behavior in different areas
- Programmatic alerting of response units (police, ambulance, tow trucks/maintenance teams) after detecting an incident
- Public transport detection across the city to monitor on-time performance and implement adaptive controls (e.g. priority passage)
- Origin–destinations traffic analysis to develop better traffic management plans and update controls in line with the most common journeys
2. Advanced safety and pollution analytics
A smart traffic management system can do more than just flash the green light on time. It can also help to design greener and safer urban environments.
Such solutions can help urban planners reach bold carbon-neutral transportation targets faster by supplying them with real-time impact data such as:
- Air quality/pollution in the area
- CO2 emissions per journey
- Traffic throughput and speeds during different weather conditions
- Road infrastructure damage post-hurricane, flooding, etc.
- Asset performance under severe conditions — heat waves or icing
- Dangerous driving behavior such as harsh braking or excessive acceleration
The above data can be collected via sensors and pre-processed on edge devices. Then it can be dissipated to a cloud-based traffic center for further analysis. Based on the obtained intelligence, you can issue better policies and controls to improve the sustainability of transportation.
Copenhagen is already testing a similar solution as part of its mission to become the first CO2-neutral European city by 2025. Local authorities have reprogrammed their traffic management system to prioritize public transport during rush hour. They are also testing dynamic traffic light programming to reduce the number of idle, fuming cars at intersections.
Copenhagen has not only cut pollution but also made travel faster — the average speed for cars increased by 4% and for buses by 9%.
Ultimately, the implementation of high-performing and sustainable traffic management systems boils down to your ability to procure real-time traffic feeds and comb through them with advanced analytics. This is a technologically complex but feasible task.
3. Predictive traffic planning
At first sight, traffic systems may seem chaotic. But an experienced manager can notice repetitive patterns:
- Regular origin–destination trips
- Problematic intersections
- Narrow, congestion-prone lanes
- Overparked streets with low throughput
- And other corners of the city where navigation gets tough
An advanced traffic management system can help you locate those troublesome areas faster and predict where traffic congestion can occur under certain conditions, such as during a heavy snowfall, after a planned event, or due to a likely road accident.
Likewise, predictive traffic planning capabilities are essential to model traffic network performance during planned events such as scheduled construction work or major public events. That’s the beauty of big data analytics in urban planning: You can estimate the capacity of your traffic network and model different response scenarios to day-to-day occurrences and unplanned events.
NoTraffic, for example, has brokered deals with a handful of cities in Arizona, Ohio, and California to install their smart traffic light systems. Using a combination of sensors, V2X connectivity, and computer vision, the NoTraffic device analyzes current road conditions. The controller can count and categorize all road users, calculate how many cars will arrive from the previous intersection, and determine how this will impact congestion at the next.
Then the traffic light dispatches data to a nearby edge device, where it’s analyzed and translated into real-time action. But traffic light synchronization happens in the cloud, allowing the entire grid to react to real-time road conditions.
If you’re interested in engaging in more long-term planning, you can pair collected traffic insights with GIS tools and mapping data to create 3D city models and simulate condition-based traffic flows. For example, how will new bridge construction impact traffic flows? You can learn more about such solutions from our case study.
4. Smart junction management
Road junctions are the biggest locus of pressure in cities because that’s where congestion and accidents frequently happen.
In Sweden, 25% of accidents on rural roads and over 50% of accidents on urban roads happen on junctions. Many of these accidents are serious, and a high proportion of their casualties are pedestrians and cyclists.
Once again, a combination of sensing technology and AI algorithms for transportation can help you make intersections safer.
To accomplish that, add the following controls to your intelligent traffic systems:
- Turning movement counts on intersections. This data can help you better interpret traffic flows and optimize signals. Plus, it can help you figure out when and why accidents occur and design alternative controls to minimize the temptation of rule-breaking.
- Dynamic traffic light signal optimization. With numbers in place, you can implement dynamic controls during rush hours and seasonal events, and you can enable users to program custom controls in line with city rules or safety planning decisions.
Case in point: Vivacity Labs helped Manchester city design and deploy an AI-controlled smart junction system. Using sensors and cameras, Manchester authorities can anonymously identify different types of road users and optimize traffic signals at junctions. They can choose to prioritize different transport modes — public transport, bikes, or pedestrians — as needed. Also, the obtained data can be reused to run traffic simulations on a range of junction layouts.
5. Electronic road pricing and toll collection
Traffic flow optimization is essential to minimizing congestion. But this method will fall short if the number of single-occupancy vehicles on the roads keeps growing.
So far, urban planners have come up with two strategies for reducing car counts on the streets:
- Entice more people to public transport by designing a better MaaS transit experience
- Put a higher price tag on cruising busy streets in a private car
The latter is called electronic road pricing (ERP). It’s a popular second-level strategy to convince people to use readily available public transport over private cars.
An ERP system relies on road infrastructure (cameras, gates, gantry systems) and in-car hardware (separate devices or onboard computers) to identify and bill cars entering a certain city area.
Advanced versions are connected to an intelligent traffic management system, which issues dynamic prices for different areas based on the time of day, vehicle size, traffic congestion, and other factors.
Singapore launched its ERP scheme back in the early 2010s. The local ERP consists of city-wide gantries on roads leading to Singapore’s Central Area, as well as on other busy roads. Using sensors and cameras, gantries capture the license plate numbers of all entering vehicles and dispatch bills to car-installed units. Drivers can link a card to the device to get billed automatically.
ERP controls are dynamic, and the pricing is designed to discourage road use during busy times. Using this system, Singapore can keep traffic moving at the optimal speed of 20 to 30 km/h on arterial roads and 45 to 65 km/h on expressways.
Hong Kong has also been doing ERP pilots since 2015 in central and adjacent areas. So are many European cities. Such projects initially create public tension. But the gains in terms of reduced travel time, less congestion, and lower pollution prompt authorities to act.
Traffic effects of different congestion charging schemes
| Traffic volume | Travel times | Public transit ridership | |
|---|---|---|---|
| London | – 16% (2006)
– 30% chargable vehicles +25% buses +15%taxis +49% bicycles -21% (2002-2008) |
-30% delays | +18% |
| Singapore | -44% after ALS
-10%-15% after ERP compared to ALS -20%-20% for other extensions of the system |
speed criteria charge levels between 20-30 kph and 45-65 kph | n.a. |
| Stockholm | -20% across the cordon | -33% delays | +5% |
| Milan | -34% (-49% for users of heavy polluting vehicles) | -17% in congestion
+7% bus speed +4,7% tram speed |
n.a. |
| Gothenburg | -10% across cordon
-2,5% vehicle-km |
-10%-20% reduction median travel time on corridors | +6% |
| Rome | -20% over cordon
+15% motorcycles |
+4% in speeds
+5% speeds PT |
+5% |
Source: Think Asia — Introduction to Congestion Charging
6. Smart parking integration
You can’t beat traffic congestion without addressing parking management.
Distressed drivers looking for parking can account for 15% of traffic in the best case scenario and up to 74% in the worst case scenario. They are also more chaotic and neurotic (if there’s no parking space) and prone to bad behavior.
The best way to get them off the street (literally) is to effectively route them to the nearest parking spot. That’s another feature you can pack into a smart traffic management system.
Apart from improving the parking experience (and collecting extra revenue), linking parking management with a traffic management system helps you collect extra data for urban planning.
For example, Ubiwhere in Portugal recently extended its smart parking management platform with urban planning functionality. Using historical parking data and the CARMA modeling language, the team came up with a solution to simulate the use of a city’s roads and parking spaces. Using this tool, urban planners can predict which areas are more prone to parking issues (especially during big events) and design better policies.
To conclude
Traffic systems are chaotic by nature. The more data we can collect about the forces shaping those flows — from weather to pedestrians — the more accurately we can predict and manage traffic conditions.
To get a comprehensive view into a system as complex as a growing city, however, a single technology isn’t enough. Truly, intelligent transportation management systems require an ecosystem of connectivity, hardware, and software technologies. And as we well know, an ecosystem is always open to onboarding new partners
Do you have a product idea for the transportation industry? Contact Intellias to learn how we can help you bring it to the market.
