Transforming rush-hour chaos: The role of connected autonomous vehicles

Imagine a city where traffic lights, vehicles, and drivers communicate seamlessly, turning rush-hour chaos into an efficient flow of vehicles.

Thanks to pioneering research from Monash University, Malaysia, this vision is closer to becoming a reality. The research team, led by Dr Susilawati from the School of Engineering, has developed a groundbreaking traffic simulation model to address the growing complexity of urban traffic.

This innovation brings together human-driven vehicles (HDVs), connected vehicles (CVs), and connected autonomous vehicles (CAVs) on signalised arterial roads, setting a benchmark in transportation technology.

Urban traffic remains one of the most pressing global challenges, contributing to time loss, air pollution, and energy wastage.

The new model focuses on the intricate interactions between HDVs, CVs, and CAVs to improve traffic conditions and reduce delays.

By simulating vehicle-to-vehicle (V2V) communication, the model allows vehicles to exchange information such as speed and acceleration, helping CVs and CAVs coordinate with HDVs effectively. This results in safer roads and smoother traffic flow.

Read more: The case for coding automated vehicles with human values

Central to this model is the “car-following” concept, which describes how vehicles adjust their speed and distance based on the vehicle ahead.

If you’re driving on a highway during peak traffic and the car in front of you slows down suddenly, you instinctively reduce your speed to maintain a safe distance. If it speeds up, you accelerate to keep up. Adjusting your speed based on the vehicle ahead is similar to longitudinal behaviour in traffic modelling, ensuring safety and efficiency in car-following scenarios.

The researchers enhanced this model by introducing a “compliance factor”, measuring how well CVs follow traffic rules.

Picture a CV approaching a traffic light. If its compliance factor is high, the CV precisely follows traffic signals and speed limits, braking smoothly as the light turns red and accelerating only when it turns green.

This orderly behaviour ensures consistent traffic flow, preventing sudden stops that could ripple back and cause congestion.

Conversely, a low compliance factor might mean the CV doesn’t fully adhere to traffic rules, potentially hesitating or making abrupt stops, disrupting traffic flow and increasing delays for others on the road.

The use of prospect theory

Using prospect theory, the research team examined driver decision-making in risky conditions, which allowed them to realistically simulate various scenarios.

Prospect theory, developed by psychologists Daniel Kahneman and Amos Tversky, explains how people make decisions when faced with uncertainty. Unlike traditional models that assume individuals always act rationally to maximise benefits, prospect theory accounts for human tendencies such as loss aversion – where people fear losses more than they value equivalent gains.

For instance, in the traffic context, a connected vehicle’s decision to brake or accelerate might depend on how the driver perceives potential risks or rewards.

By incorporating this framework, the researchers could model how CVs and their drivers behave in high-stakes traffic scenarios, such as avoiding collisions or navigating congestion.

Read more: Will self-driving cars solve the problem of traffic congestion?

The model’s ability to integrate HDVs, CVs, and CAVs into a cohesive system is a leap forward. Results from simulations showed that increasing the penetration rate of CVs and CAVs led to significant improvements in traffic performance.

As the penetration rate of either CVs or CAVs increased, queuing delay length decreased, affirming the hypothesis.

However, the average arithmetic speed of mixed traffic flow also declined, suggesting that more conservative speeds contributed to reduced queuing, ultimately saving time and fuel.

The highest queuing delay recorded in the network was below 10 seconds for simulations involving 100% HDVs.

Notably, higher queuing delays were observed when CVs and CAVs made up 30% of the traffic mix.

In contrast, other scenarios saw significantly lower delays – an insight with real-world implications for urban traffic management.

A challenging scenario

Creating this sophisticated model was no small feat. One major challenge was the limited availability of open-source data on CAVs. The team also had to reconcile the unpredictable behaviour of HDVs with the precision of CVs and CAVs, ensuring the system remained realistic and functional.

They addressed these challenges by calibrating the model to handle diverse road designs, signal timings, and driver behaviours.

This research aligns with several UN Sustainable Development Goals (SDGs), including Goal 11 (Sustainable Cities and Communities) and Goal 13 (Climate Action).

By enabling smoother traffic flow, the model has the potential to cut greenhouse gas emissions and reduce energy consumption, contributing to cleaner and more sustainable urban environments.

The traffic model also supports Monash University’s Impact 2030 strategic plans. By addressing climate change through reduced emissions and energy use, the model helps contribute to Monash’s mission to drive global environmental sustainability.

Its emphasis on safer and more efficient urban mobility aligns with the University’s commitment to more thriving and connected communities.

Read more: Autonomous cars are the future of transport? Don't believe the hype

The researchers are already looking to further enhance the model. Plans include incorporating various intersection designs and urban layouts, as well as integrating predictive analytics to anticipate and prevent traffic disruptions.

A real-time feedback system is also under development, which could dynamically adjust traffic flow based on current conditions.

The team is expanding the model to include environmental metrics such as emissions and energy consumption to align even more closely with the UN’s SDGs. These metrics will be evaluated by integrating the traffic simulation model with the latest emission model, a feature available in the newest version of PTV VISSIM.

These enhancements aim to quantify the environmental benefits of increased CV and CAV adoption, reinforcing the case for transitioning to intelligent transportation systems.

Partnerships prove pivotal

Monash University, Malaysia provided essential resources, including access to PTV VISSIM simulation software and high-performance computing infrastructure. The University's focus on intelligent transportation systems (ITS) and partnerships with industry and government stakeholders were pivotal in advancing this research.

Monash’s support has been instrumental in overcoming challenges and ensuring the success of this model. This includes access to the latest version of PTV VISSIM traffic simulation software, and high-performance computing resources to handle the computational challenges of large CAV and CV trajectory data.

As urban centres worldwide strive to meet sustainability goals, innovations such as this traffic model are crucial. By refining this model and implementing it globally, cities can create safer, greener, and more efficient transportation systems, fulfilling their commitments to sustainable development.