Like spider webs, economic supply chains weave through our interconnected world – efficient links between suppliers and manufacturers around the globe, yet also vulnerable to disconnection. The force of a typhoon in the Philippines or Japan sends shockwaves quickly along the global delivery network. Forecasts that climate change could bring more intense storms and more frequent heavy rainfall and drought (IPCC Special Report) raise the need for new ways to analyse the risks of potential interruptions in the manufacture of everything from mobile phone chips and car brake components to anti-HIV drugs.
Typically, risk management builds upon the assessment of how likely a disruption at a certain node of the network would be and how much it would cost. The impact of an average supply chain interruption, like a transportation breakdown, is calculated using historical data.
But there is neither historical data nor a precise way to forecast natural catastrophes and extreme weather events that could result from climate change.
“Classical risk management techniques are not appropriate when it comes to assessing low-probability risks like natural disasters, environmental challenges, geopolitical problems, epidemics,” said David Simchi-Levi, professor of global operations in MIT’s department of civil and environmental engineering.
“Each of them happens infrequently, but there are so many types of sources of risks that you don’t know what will reach the supply chain the next 12 days or 12 months, but it is sure that something will go wrong,” Simchi-Levi said in an interview with Road to Paris.
To face that challenge, scientists turn to mathematical models to identify the weak spots of all the webs – independent from their direct economic importance – and show how the networks are interconnected.
Simchi-Levi developed a model that analyses where the real risks in a supply chain are – leaving out what classical risk management focuses on: the probability of a risk. Instead he decided to focus on the economic impact of a failure: “I measure risks with dollars. I estimate the cost of risk and this provides an ability to understand what will be the impact on the bottom line if something goes wrong.”
In his analyses of companies’ supply chains, he found that the highest risks often lay where the risk managers did not look. “Most companies will equate high risk with supplier sites where total amount spent is very high.” But he found that was not where the real risks lie. The real risks were hidden in overlooked supplier sites with relatively low amounts of money spent.
“The problem with risk is that it is almost a philosophical challenge: if a company’s executive could cut costs by five percent or increase revenue by seven percent, that executive would be made a hero. But suppose you are in charge of risk mitigation strategy, and because you are successful nothing happened. It’s hard for people to understand the importance of ‘nothing happened’. Nobody is rewarded. So, sometimes it is hard to build incentives for companies to focus on this,” said Simchi-Levi.
“The market needs an educational process to understand this. More and more companies are starting to think about this. Because they saw in 2011 with the tsunami in Japan and the flood in Thailand that supply chain disruptions can have a huge impact – not only on the supply itself, but more importantly on stock performance. That is why you have more CEOs and CFOs talking about this new way of risk assessment in supply chains.”
While companies run their supply chain analysis behind closed doors, a new project in Potsdam, Germany, zeean, chose a rather unusual way to bring the global network into the light – relying on public input to understand economic flows around the globe. They hope their approach will help transform supply chains into less vulnerable and more resilient networks.
“It’s about understanding the economic network of the world – which we hope that people will find exciting so that they start searching for and entering data and thereby become part of this whole endeavour,” explained physicist Anders Levermann, who is leading the zeean project.
The project’s website, zeean.net, was launched in February 2014. The idea is similar to Wikipedia: anyone can register as a user and enter data that is then cross-checked by other vetted users or their team of mathematicians and physicists.
Based on that economic data, Levermann and his team set up a network analysis model showing what happens when a country drops out of the chain. “With regard to Typhoon Haiyan, which hit the Philippines in 2013, we computed which impact a failure of the Philippine’s economy would have. This country exports half of the world’s coconut oil, which is used in food production worldwide. A breakdown would affect, for example, six percent of U.S. production directly, and even 21 percent indirectly,” said Levermann, who is also Chair of Sustainable Solutions and Head of Global Adaptation Strategies at the Potsdam Institute for Climate Impact Research.
With this method, called hindcasting, the scientists tested their model against real events. “We want to do this not only with climatic events, but also with disasters like Fukushima and see whether we can understand the observed economic repercussions of these events in the statistical data,” he said.
“That is what you can compute on the zeean website right now: if you ‘knock out’ one country, you can see how that influences economies of other countries,” he said.
Yet, with only 70 contributors so far, the data is too coarse to provide the insight needed for companies or insurers to make economic decisions. The aim is to go into more detail, both spatial and in terms of economic sectors or goods. “The data for downscaling our model are widely available for, for example, the départements in France, the provinces in China, the Bundesländer in Germany, and [each of the] states in the U.S.,” he said.
Zeean would allow everyone – from the average person to policy makers and companies – to refine the model according to the place or sector they are most influenced in by entering data for that specific spot – and still see how this influences, and is influenced by, the global operations network.
“On zeean we want to gather all the data with the help of the crowd out there,” Levermann said of the citizen-science project. “And then, whenever someone enters new data, we will adjust the network accordingly, so it is consistent with this new data.”
Climate risk assessment, supplemented with zeean, would show which countries are rather safe production sites in terms of likeliness of climate change impacts – and which are rather risky. That could mean economic disadvantages for the latter. “But that is exactly why it is so important that the data is open to the public – so everyone can adjust to it,” Levermann said. “I see that a problematic situation might emerge, but I think that hiding the truth is not going to help anyone either.”