Physics of Risk is becoming more and more popular and gaining more attention from the media. This time our permanent contributor Aleksejus Kononovicius gave an interview to a journalist Marija Rudzevičiūtė from Bzn start. The original article was published in Lithuanian language, but here you can find its free form translation to English.
"Probably most of us were stuck in a traffic jam for at least once in our life. Some of us were certainly stuck in a traffic jam, though there were no car crash, no road works or no other apparent reason for the road to be jammed. Yet all cars move in a jerky motion - they suddenly accelerate, move a bit and then suddenly stop. The reason lies in a small errors made by the drivers themselves", everyday situation in a large city is reviewed by Aleksejus Kononovicius, younger research fellow at Institute of Theoretical Physics and Astronomy, Vilnius University. He is also a current contributor to the "Physics of Risk", a website which presents an idea that "more physics - less risk" for a broader society.
Small disturbances - large consequences
According to our interviewee, if road is packed with cars, then sudden
deceleration of one car may trigger "stopping" wave through all cars in
"By taking these observations into account we can create agent-based model, which would replicate general features of car traffic. The model could be further applied while creating, testing and improving traffic related projects - cross-roads and flyovers, self-driving cars and smart traffic lights. These tools would be certainly used by governmental institutions and architects, but the benefits would reaped by the society", - young scientist suggests one way of improving our everyday life by relying on physical thought.
"In the second half of the XX-th century physicists noted that all the complex structures, such as a life and culture, emerged from a simple initial conditions, few elementary particles, during the Big Bang. It was somewhat peculiar that no one has thought of that before and that no one asked a very natural question - how can complexity emerge from simple conditions? Some of the physical systems as well as biological, social and economic are far from equilibrium", - replied A. Kononovicius when asked about the relationship between physics and every day life. Though he also noted that every system is unique and one should not neglect it.
More physics - less risk
"It is our motto", - explains the young researcher. - "It is a nice play with words, which reflects actual scientific research. The better the understanding of phenomenon, the easier it is to estimate and control the related risk. This is especially important in the context of social and economic systems. But this problem concerns not only physicists, many excellent works were also published by mathematicians, economists and sociologists", - the relation between physics and risk is unveiled by the interviewee.
Asked if this motto could applied in business A. Kononovicius swiftly replies: "Business, as well as society, would gain a lot from the optimization of logistic problems. One of the success stories in business world is related to the American "Southwest Airlines". This company was one of the first to use agent-based model to optimize its cargo delivery policies. Agent-based model suggest counter-intuitive strategy - to decrease not the distance traveled, but number of flights. This strategy helped save millions dollars, which would be otherwise spent on cargo handling."
Biggest risk in the financial markets
According to the interviewee financial markets is the business sector related to the biggest risk. "Here we have to face lots of traders optimizing their risk and expected returns every day. Currently, as far as I know, they use phenomenological models, which are either derived from the past data or are fitted to it. The essential drawback of such models is that they are able to reproduce only what is included in the data - they are able to grasp only the most direct relationships between the observables. Agent-based models, on the other hand, may grasp the hidden relationships and thus replicate the full complexity of the phenomenon, both one that exists in the data and the one that may emerge", - A. Kononovicius develops his thought on the financial markets.
According to him, agent-based models might grasp the hidden feedback loops, which exist in the actual systems. The inclusion of feedback loops into the models may help to understand the impact of our present actions on the future of the market. This understanding might provide a way to make smarter decisions, which would also anticipate the future developments in the market. This should work very well on the short time scales.
"Currently there are numerous commercial proprietary software packages, which allow anyone to model business processes on almost any computer", - A. Kononovicius comments on the current situation and the possibility to do the simulation at home or at the office.