Evacuation Modelling: Finding the Best Time (And Way) To Get Going (Op-Ed)
This article was originally published at The Conversation. The publication contributed the article to Live Science's Expert Voices: Op-Ed & Insights.
Reports from the Philippines reveal a lack of typhoon preparation and evacuation efforts.
When to evacuate – and how – spells the difference between life and death. As we know, typhoons can cause widespread flooding of surrounding areas, and don’t just affect what lies in the path of the storm. Planning an evacuation is a game against nature.
Typhoon Haiyan (and similar events around the world) indicate that people do not play this game well … but computers do.
In collaboration with the Los Alamos National Laboratories and the US Department of Homeland Security, our team of researchers at NICTA has developed algorithms to mitigate the impact of natural disasters. They were, for instance, activated to help in power restoration during recent massive Hurricanes Irene and Sandy.
So can we learn from international disasters, and can we apply that information to the Australian context? The answer is a resounding yes to both.
Our team focuses on planning large-scale evacuations for Australian cities. Its goals are twofold:
Get the world’s most fascinating discoveries delivered straight to your inbox.
- to inform policy on what constitutes a good evacuation plan
- to demonstrate that computers are game-changers.
Planning an evacuation is an extremely challenging task, with calculating the best procedure akin to finding a needle in a haystack. There is an astronomic number of possible evacuation plans to consider, even for a small city.
Although few plans are safe, the number and complexity of decisions quickly becomes overwhelming – especially as rising water or traffic accidents block roads – but computers can dramatically help emergency services design evacuation plans which people can actually follow.
To be efficient, computer tools must understand congestion and human behaviour. They also need to take into account new information that will become available in real time via river gauges, traffic monitoring or social media.
Ultimately, they can be used to replan part of the evacuation in real time. Let’s look at an example.
An evacuation scenario
The evacuation tools developed at NICTA not only push the frontier of emergency decision support systems – they also show that optimisation algorithms have the potential to save lives during disasters.
Consider, for instance, the case of the Hawkesbury Nepean region in NSW. Experts estimate that a flood of the magnitude of the historic event of 1867 will cost A$2.5billion and affect thousands of residents.
Such an event would see the water reach 19.2m at the Windsor bridge, compared to the normal level of 1.5m.
As a result, 200km2 of surrounding land would be inundated with fast-moving water, threatening the lives of about 70,000 residents. An increasing population, combined with the silent threat of the Warragamba Dam spilling or failing, make this region one of the worst flood risks in Australia.
To evaluate the impact of major flooding on regional areas, we developed a tool to simulate the evolution of the flood, minute by minute. The algorithm forecasts a 24-hour flood in a few minutes and specifies which areas need to be evacuated and at what time, and when roads will be flooded.
The optimisation algorithm then uses all this information to produce an evacuation plan indicating when each person needs to evacuate, where he or she should go, and route to follow.
Contrary to human decision makers, the optimisation algorithm considers simultaneously all 70,000 evacuees, how they interact and compete for road capacities, and how the flood affects the transportation network. The whole plan is produced in a matter of seconds and rescheduling can be performed in real time as the event unfolds.
The video above illustrates the evacuation of 70,000 persons in the Hawkesbury Nepean area. The flood (in blue), rising from the Warragamba dam (at the back) inundates the flood plain. Vehicles (in green) are evacuated following precise evacuation routes and schedules to shelters (green boxes).
In this illustration, emergency services wait as long as possible before giving evacuation orders in order to avoid false alerts.
Lessons to learn
Our algorithms have led to fundamental insights about evacuations and human behaviour. Letting every individual decide when, where and how to evacuate can have disastrous consequences.
In the Hawkesbury Nepean flood scenario, if everyone leaves at a reasonable time and goes to the closest evacuation shelter, more than 60% of the population will not reach safety and will end up trapped by major traffic jams.
It is not surprising that independent decisions by 70,000 individuals do not lead to an effective evacuation – but our algorithms evacuate every single person.
Even better, if as many as half of the population does not follow the plan exactly but leaves at the time it is instructed to, 97% of evacuees will reach safety.
Equally interesting is the fact that most existing evacuation algorithms are too optimistic: they delay the evacuation too much and, as a result, a substantial portion of the population cannot be evacuated.
After the 2005 Hurricane Katrina, the US recognised the need to go beyond situational awareness and to adapt for disaster management the optimisation algorithms used in airlines, logistic systems, and supply chains.
Technological advances may have significant impact, and save numerous lives, in Australia and the Asia-Pacific region. The next step, of course, is convincing more governments around the world to deploy them.
Further reading: A Conflict-Based Path-Generation Heuristic for Evacuation Planning
Victor Pillac works for NICTA.
Pascal Van Hentenryck is the leader of the Optimisation Research Group at NICTA.
This article was originally published at The Conversation. Read the original article. The views expressed are those of the author and do not necessarily reflect the views of the publisher. This version of the article was originally published on Live Science.