by Professor Majid Sarvi, Transport for Smart Cities, The University of Melbourne
Driverless vehicles seem like they could belong in a distant world, but in reality, there is no doubt that we are moving towards a future where vehicles will be self-driving.
We know that our roads are congested and dangerous, and many would argue that the burgeoning industry around autonomous vehicles is the way forward. By taking out human error and recklessness, these vehicles have the potential to make our roads safer than they have been since automobiles first appeared in the early 20th century.
However, autonomous vehicles in and of themselves are not the only answer. To truly mitigate the risks of driverless cars and to seize the opportunities offered by these new technologies, we need to implement systems where all elements of the environment are communicating and reacting to one another.
This is what the University of Melbourne Transport Technologies program aims to do. We bring together the whole network, creating an ecosystem that will enable cities to deliver a smarter, safer, more efficient and more sustainable transport future.
Our key demonstration of this aim is Australian Integrated Multimodal EcoSystem (AIMES), which I founded in 2016 to bring together public and private sector domestic and international partners.
A connected environment
Using sensors across all elements of the built environment, we can connect everything from public transport, to cars, bikes, traffic lights, intersections, pedestrian movement and even footpaths. This allows us to see all road users, not just those in vehicles. We are no longer privileging cars alone, but looking at how all transport system users interact with one another.
Through distributed cooperation, we have the ability to monitor traffic congestion, weather conditions and hazards in real time, and to adjust accordingly. This has the potential to save time by diverting traffic away from congested spots and also to save lives. Such monitoring not only provides a picture of what is happening, but when used in conjunction with autonomous vehicles, can allow for machine intervention.
What also makes this exciting, is that this type of technology has the ability to evolve. It is ‘intelligent connectivity’ that has great potential to adapt as our cities grow and our infrastructure changes.
So far, we have seen enormous advancements in autonomous vehicles. But these work well in controlled test environments, and are less able to navigate the constantly changing and variable conditions of real-world traffic. This makes the argument for integrated transport all the more compelling.
The University of Melbourne, along with its 50 partners through AIMES, is working to address this challenge with a world-first solution. By integrating sensors into the CBD-adjacent AIMES ecosystem environment and into autonomous vehicles themselves, we are monitoring in real time the flow of vehicles, cyclists, pedestrians and public transport through the grid.
This world-first sees the largest ever inner-city grid of streets mapped with smart sensors. We have recently completed phase one of testing and will be using results to implement further changes to the ecosystem, including enhancing the technology and expanding the grid.
The benefits extend beyond safety, to allow congestion minimisation and more sustainable practices. By better integrating and connecting our road systems, traffic can be diverted down different routes, allowing for more efficient use of our current road infrastructure. This integrated, connected and multimodal approach is particularly important as we compete with burgeoning population growth and increased pressure on city mobility.
An integrated system doesn’t have to solely monitor road users. It can also incorporate environmental conditions, including air quality information. Smart sensors will be able to feed information to drivers to alert them when conditions change. This could mean alerting a car to slow down during wet or windy weather. Given Melbourne’s infamously unpredictable weather, this is another important factor to consider when looking for ways to create a safer driving environment.
Connected intersection trial
We recently trialled four different scenarios at our ecosystem in Carlton, Melbourne. Aiming to improve the response times between road sensors and a connected autonomous vehicle (CAV), our real-world test showed the potential that comes with leveraging edge computing infrastructure along with CAV capabilities.
The tests involved real-life use cases including speed management, intersection collision avoidance and vulnerable road user protection. Without relying on human intervention, sensors in the ecosystem were able to prevent potential accidents by issuing early warnings. Delivering this information at speed is vital, with a two-second delay to a vehicle moving at 50km/h (urban speed limit) equating to an extra 48m of travel.
The trial was conducted in association with key AIMES partners including Cisco, VicRoads, Cohda Wireless, WSP and the Victorian Government.
It was the first time globally an Internet of Things (IoT) edge and fog computing solution ran on a roadside intersection, using analytics technology from Cohda Wireless over which Cisco Kinetic IoT Data Fabric had been built.
The test environment in inner-suburban Drummond St, Carlton, is known for its popularity with pedestrian and cyclists. It offers a complex urban environment and also the ideal ecosystem for a living laboratory, just a five-minute walk from the University of Melbourne’s Parkville campus.
The AIMES Bus
In March 2018, we launched a partnership with EasyMile and now have our first autonomous bus on campus. This mini bus is ideal for the low-speed testbed environment and is fitted with open-platform technology, meaning that it can be updated as software progresses.
As researchers, the most important thing is investigating the transport ecosystem and using the insights to understand how every passenger, pedestrian or cyclist moves through a city. Our work is about efficiency, safety and sustainability, not about privileging one form of user over another.
As everyday objects are becoming increasingly sophisticated, and with the IoT, objects that can perceive and respond to their surroundings will become the norm. Not only will AIMES help achieve communication and eventually cooperation between road users, it will help us as researchers to learn, reshape and ultimately influence the lives of every day road users.