Computer model tackles supply chain disruptions

Using mobile phone data, a team at CSH mapped an entire nation’s supply chain network. The computer model, which predicts what’s at stake if supply chains break down, can be easily implemented by other countries


Researchers at CSH used simple telecommunications data to map an entire country’s production network, including all relevant firms and supply relationships with their clients. The computer model, based on mobile phone data, was able to predict the systemic risk of each company within the country and estimate their resilience.


“Our model is cheap, scalable, and easy to implement by any country that has telecommunications data available. It allows leaders to prepare their economies for future shocks that have the capacity to inflict substantial disruption in their supply chains,” explains CSH scientist Tobias Reisch, one of the co-authors of the study published in the journal Scientific Reports available online June 3rd. 


Additionally, the new method provides an image of the country’s entire economic behavior on a timescale of a single day, highlights Stefan Thurner, CSH president and co-author of the study. 


“This precision enables monitoring of short-term changes in the economy, such as the effects of Chinese Covid-19 lockdowns on European companies from day to day. A standard economic data set can’t provide this kind of information due to delays of several months or even years,” says Thurner.  “We might be able to use this new method to strategically increase the resilience of the economy under shocks and crises. And it comes at practically zero cost.”


The team reconstructed the supply chain network of a medium-sized European country based on mobile phone data provided by a telecommunications service provider. The dataset – which includes various attributes of the call, such as time, duration, completion status, source number, and destination number – corresponds to tens of thousands of companies. The data, which was collected over a period of four months in 2020, is classified as highly sensitive and is, therefore, anonymous. 


 In their analysis, the researchers also included additional information on the firms’ industry classification and balance sheet information. And they found that, when there was a strong communication link between two companies, the probability of observing a supply link was about 90 percent. “If two firms would talk on the phone for at least 5 minutes per week, we could establish the flow of goods and services between them,” explains Reisch.


By analyzing the communication networks, the team was able to reconstruct the national production processes and supply relationships between companies in an unprecedented way. Then they calculated how supply chain disruptions affected each enterprise.


According to the study, about 65 companies have the potential to affect large parts of the economic activity. These enterprises display an extremely high systemic risk of about 21 percent – meaning that,  if firms can’t adjust their supply relations, about 21 percent of the national production could be adversely affected. Apart from these core firms, the systemic risk of companies is generally small. 



The results indicate that, in terms of systemic risk, firm size is not a good predictor. “We found both small and large firms within the 65 top high-risk firms group,” says Reisch. The majority of the top high-risk companies operated in the manufacturing sector, followed by those in the electricity, gas stream and air conditioning supply, and financial and insurance activities sectors.


 The new paper supports the findings of a previous CSH study that showed how a few enterprises connected by a network of highly critical supply chain relationships can shake up the economy. In that work, CSH scientists used value-added tax (VAT) information – a general tax that, in principle, is applied to all goods and services – to reconstruct the supply chain networks of the Hungarian economy. 


Reisch notes that granular data on a nation’s supply network, such as VAT data, are notoriously difficult to acquire. “Our approach is one of the very few alternative ways of getting a fine-grained and comprehensive view of national supply networks when there is no VAT available,” adds Reisch.


Most countries have little knowledge of how their supply chains actually look like, let alone how they change in response to disruptions, points out Peter Klimek, CSH scientist and co-author of the paper. “Our approach provides governments and policymakers with comprehensive and actionable information about emerging and critical vulnerabilities in the supply network, allowing for timely mitigation measures to increase production network resilience,” evaluates Klimek.


 The study Monitoring supply networks from mobile phone data for estimating the systemic risk of an economy, by T. Reisch, G. Heiler, C. Diem, P. Klimek, S. Thurner, was published in Scientific Reports 12 (13347) (2022).




T. Reisch, G. Heiler, C. Diem, P. Klimek, S. Thurner,
Scientific reports
0 Pages 0 Press 0 News 0 Events 0 Projects 0 Publications 0 Person 0 Visualisation 0 Art


CSH Newsletter

Choose your preference
Data Protection*