Printer friendly version
The anti-crisis project: Thinking our way to a better future
02 August 2010
Just a few days ago, at the Love Parade dance and music festival in Duisburg, Germany, 21 people were trampled to death in a human stampede. Apparently, the responsible decision-makers did not see the problem looming before the event. But couldn’t we foresee at least some of the disasters before they strike?
In fact, scientists say, with the right kind of effort, we could do much better, not only with crowd disasters, but also in managing traffic problems, impending energy shortages, persisting financial and economic instabilities, climate change, and many other challenges. An ambitious new scientific effort now aims to transform our ability to understand and manage mutually entangled social, economic, and technological systems through an Apollo-like project uniting scientists across Europe. The FuturICT initiative will exploit the revolution in modern computation and information technologies to build an immeasurably more powerful science of human systems and their interaction with the global environment.
The project is being coordinated by a team of scientists led by physicist, traffic scientist, and sociologist Dirk Helbing of ETH Zurich. "It’s time to explore social life on Earth, and everything it relates to, in the same ambitious way that we have spent the last century or more exploring our physical world," he says. Their proposal, submitted for the European Commission's Flagship Programme, aims to assemble expertise across the whole spectrum of science—from physics, computer science, environmental science and economics through psychology, ecology and sociology—and, by developing supercomputing facilities and large-scale laboratories, to build a much more powerful human science on which to base future policies.
"This may well be one of the most profound scientific initiatives of the 21st century," says Jeffrey Johnson, President of the European Complex Systems Society.
The information revolution facilitates a better understanding of society
In the past decade or so, modern information and communication technologies—especially through the Internet, World Wide Web and mobile communications—has begun transforming social science into a far more quantitative, and even predictive, science. Social scientists have also learned to augment the power of their minds with computing technology, building models of social systems—a community, a company, a market, etc.—in which computer agents act like people, make decisions as people do, learn and adapt. With funding of approximately EUR 100 million per year for ten years, the FuturIcT project aims to build on these advances to establish three systems on an unprecedented scale:
A Living Earth Visualator. A super-computing platform capable of simulating and visualizing the world at the global scale, focusing on the interaction between technical, social and economic systems as well as their physical and biological environment. This will provide a setting for the exploration of whole-Earth dynamics and policies designed to manage them.
Crisis Observatories. Laboratories running massive data mining and computing systems to detect possible crises, such as bubbles or crashes in financial or housing markets, to gain advance warning of critical shortages in, say, oil, water, or food, or to develop ways to identify risks of wars and social unrest, disease spreading, or environmental instabilities.
Knowledge Accelerator: A concept bringing together key representatives from a wide range of scientific disciplines, from businesses and other organizations, and from governments to identify important social or technological innovations early on and to devise ambitious, practical programmes to further their wide social benefit.
The project envisions these three components working together to move from the early identification of a problem or opportunity through its analysis to systemic designs and the visualization of the impact of policy measures by means of Decision Arenas. For example, a Crisis Observatory might gather and analyse data from stock markets, banks and other businesses, inflation, the housing and labour markets, resource availability and so on, and carry out massive real-time data mining and computational simulation to identify advance warning signs of potential financial or economic crises. Such information would go to experts and decision-makers who would further investigate the potential problem in large-scale computer simulations in the Living Earth Visualator and explore the likely effectiveness - and also the possible unintended consequences -- of proposed policies. In the case of a looming market crash, for example, they might test the likely consequences of a ban on naked "short-selling" or other speculative trading practices, or anticipate how market function would likely change, if authorities stepped in to limit leverage. Computer experiments in the simulated economy might discover effective but non-obvious policy steps that no individual would ever have imagined.
Including the human component
People, of course, don't always respond the way policy makers intend. Hence, the more promising policies identified through simulations could be further tested in massive online games in which real people would participate in the experiment. Imagine, for example, 10 million people playing within virtual worlds mimicking the international financial system and having real incentives to act realistically in their roles as governments, central banks, businesses or labour groups within Europe. Such a "parallel world" system -- akin to Second Life -- would offer an unprecedented means for foreseeing and comparing the possible outcomes of policy decisions. For example, what's the likely outcome for the European economy, if the European Union backs up Greece and the other financially troubled nations? What would change if, instead, these countries would again introduce independent currencies?
After this testing stage, the results of the different computer scenarios and parallel world experiments would be presented to decision-makers in a Decision Arena -- an information-intensive environment designed to give decision makers the ability to explore any policy matter from as many angles as possible. Based on political priorities, moral factors and the myriad other issues that influence real-world decisions, they could make their policy choices, informed by the best available insights from science.
A similar process of sensing, analysis and response, streamlined to work in real-time, could most likely have prevented the deadly stampede in Duisburg, and many others that happen every year. With modern automated video surveillance and analysis, computers monitoring a crowded location and calculating its likely future could easily detect warning signs such as a drop in the flow or a sudden increase in the crowd pressure. They could also predict the likely consequences of a gate being blocked, and give real-time warnings to police. Furthermore, they could offer advice regarding crowd management options and their likely impact on the system. Such a real-time monitoring and decision support system would be extremely useful to reduce the risk of a crowd disaster. So far the feedback loops that would bring the data and computer simulations together in real-time and inform the decision-makers are missing. The challenge here is to develop an integrative systems design that helps to manage complexity.
Although the systems FuturICT wants to build are sophisticated in amibition and scale, it is rapidly moving toward reality. For example, great progress has been made in large-scale computer simulations that will play a role within the Living Earth Visualator and the Crisis Observatories. These include simulations informing policy, for example, on traffic management in large cities, evacuation scenarios in response to external threats, or the spreading of diseases. Moreover, economists, working with a broad group of other scientists, have recently begun to develop simulation platforms for the entire European and US economies.
FuturICT aims to bring all this together on an unprecedented scale
"This is an experiment," says Joshua Epstein of John Hopkins University, "that we cannot afford NOT to do."
The FuturICT initiative aims at maximum transparency. An ethics committee will ensure that its research benefits society and that the data-mining approach will protect privacy. The project now seeks collaborations with business and industry, with government and administrative institutions, foundations and funding organizations to unleash the full potential of the project and create the societal and economic leverage effect expected by the European Union. Please contact Steven Bishoph (email@example.com), Paul Lukowicz (firstname.lastname@example.org), Josh Epstein (email@example.com), or Dirk Helbing (firstname.lastname@example.org) for further details.
The proposed project involves more than 200 scientists from over 50 universities and institutions including ETH Zurich, Cambridge University, University College London-UCL, Oxford University, Imperial College, Politecnico di Torino, Centre National de la Recherche Scientifique, EPFL, La Sapienza University of Rome, University and TU Munich, Potsdam Institute of Climate Research, Central European University, King’s College London, London Business School and Open University, to mention only a few. A wide range of science organizations and George Soros have written letters in support for the initiative.