It's not a typical story for the Embedded blog, but identifying network congestion is a key challenge around the world, particularly in airports.
Many airports are operating beyond their maximum capacity due to the growing number of passengers and cargoes. Measures are being taken to alleviate congestion, such as increasing the number of staff on ground or expanding airports. However, understanding human behaviour and the actions people take depending on their attributes and information given is a key challenge to tackle.
Professor Shingo Takahashi of the Department of Industrial and Management Systems Engineering at Waseda University and Fujitsu Laboratories developed a new technology that automatically analyses the factors leading to congestion based on the results of human behaviour simulations.
Conventionally this analysis uses the results of large numbers of congestion prediction simulations to try and find the root cause of congestion, but this overlooks potential causes due to human error from time to time and required the simulations to be evaluated one by one, which sometimes took several months.
"The new technology groups categories that have a certain degree of commonality, and expresses the characteristics of respective agents (which represents a diverse group of people) in a small number of combination categories without listing the results of movements and routes of tens or hundreds of thousands of agents individually through simulation-based modelling," said Professor Takahashi.
"This makes it easier to discover the cause of congestion and answer the question of what sort of measures can be taken to change the mindset or actions of people with specific sets of attributes in a matter of just few minutes."
Takahashi and Fujitsu applied this technology to a simulation at a local airport in Japan. They discovered approximately four times as many causes of congestion in comparison to analysis by experts.
In an analysis of congestion at security screening, the system found that passengers lining up at a specific check-in counter caused sudden congestion. The measures implemented according to the findings of the technology reduced the number of people waiting in line by one sixth than the measures proposed by experts. Additionally, the number of staff required to implement the measure was reduced by two thirds.
This technology enables a quick evaluation of measures to ameliorate congestion in commercial facilities, event venues and other locations that deal with congestion due to high attendance or urban centralisation as well.
The work has significant implications for the Internet of Things, linking to digital signage. Fujitsu will also use its Human Centric AI Zinrai artificial intelligence and supercomputer technologies, and together with its Fujitsu Technical Computing Solution Citywide Surveillance software, which enables a real-time understanding of urban conditions. These will be used to develop a future predictive solution for congestion.