Rebooting Crisis Management With Transportation Data and AI

Image credit: iStockphoto/NicoElNino

The recent crises in Australia are unlocking the hidden value of transport-related data.

Many of the digital applications based on this data were developed before the pandemic or the recent bushfires. The whole area of traffic management was getting a big reboot, while the entry of autonomous cars made new data-driven applications necessary.

Now, researchers are finding out that the same applications that combine sensing and optical technology with AI can also have an epidemiological application.

Stopping the virus spread

In Australia, the researchers at Data61, the AI and data science arm of Australia’s national science organization CSIRO, and state government agency Transport for NSW are collaborating to manage the pandemic.

At first, Data61 focused on Australia’s largest city of Sydney to develop a Traffic Assignment Engine program for “analyzing automated end to end, multi-modal journey planning for operators and passengers.” It had been in prototype development for a few years.

The Engine came into its own with a rollout during the COVID-19 disruptions, a time when understanding congestion was vital to improving social distancing and minimizing the spread of the coronavirus.

Social distancing, here, is not just where you stand in a queue. It is a function of people’s flow through the city to minimize human density and the likelihood of transmission.

The efforts gained recognition at the recent annual Intelligent Transport Systems industry awards in their eleventh year. The awards recognize the increasing contribution of new technology in the transport industry.

Navigating safer transit

During COVID-19, the Traffic Assignment Engine was matched with a Disease Spread Model. This effort was claimed to be the first in the world that combined detailed transport simulation with cutting-edge epidemiological modeling to tackle the challenge of safe transit during a pandemic.

The original AI Engine aggregated data generated by tens of thousands of sensors throughout the road network, coming from the fare-paying Opal card, GPS devices, traffic signals, and buses. It also includes newer sources such as mobile phones, which provide vehicle speeds and deliver insights on congestion. 

From running this data through the AI Engine, it is possible to model the impact of network changes or disruptions and then issue automatic journey planning information for transport operators and travelers. It develops accuracy in knowing what will happen in the future, using this knowledge to improve congestion planning.

The result is that the combined system can make accurate predictions two hours into the future and act in five minutes, a capability that is likely to lead to significant improvements in efficiency in managing traffic flows. It also effectively manages social distancing to minimize the spread of the virus.

Fighting fire with data

Data61 has also been busy in deploying data science to fight bushfires. Australia was devastated by fires in the summer of 2019 and is about to go into another fire season on high alert.

The Data61 Spark project is an award-winning open framework for bushfire prediction and analysis created by Data61.

It takes current knowledge of fire behavior and combines it with state-of-the-art simulation science to produce predictions, statistics, and visualizations of bushfire spread.

The results can also be used for land management and planning, fire mitigation analysis, real-time fire prediction, and analysis of fire events.

Spark can read weather data from meteorological forecasts and use this information directly within fire models.

Geographic information, such as land slope, vegetation, and un-burnable areas, such as roads and water bodies, also affect the spread of the fire. Spark allows users to easily incorporate such environmental data and to use this information to define a fire spread rate.

Predicting bushfire’s spread provides a range of benefits across various sectors, including infrastructure planning, land management, emergency services, and more.

Unlocking crisis insights

Neither of these data projects is a magic bullet to fighting the pandemic or bushfires. But they don’t necessarily have to be.

What they do is aggregate multiple data sources to drive understanding of patterns to improve prevention and planning.

The positive news is that AI can learn through repetition, which means the systems will improve over time, and prediction modeling will become more accurate.

In a world where pandemics and natural disasters are likely to become more prevalent, it is reassuring that now we have AI on our side to help us in what has become a life and death struggle.

Image credit: iStockphoto/NicoElNino