school Biologically-inspired information and communication technology

Biological systems are inherently scalable, adaptive, and robust. Once we can derive and adopt fundamental principles behind those characteristics, highly sustainable information and communication systems come into reality and keep playing a role as a social infrastructure. In our Bio-system Analysis Laboratory, we conduct such bio-inspired research to analyze and apply biological algorithms and mechanisms to the establishment of novel information and communication systems and technology.

Information and communication systems, similarly to other engineered and artificial systems, have been optimally designed, built, and managed to accomplish the best performance in the assumed operational conditions. They adapt to changes and recover from failures by preparing adaptation and recovery mechanisms to maintain their performance. However, ever-increasing size, complexity, heterogeneity, and dynamic changes make such conventional methodology infeasible. Information and communication systems often face unexpected events and unassumed conditions and as a result drastic performance degradation and severe halts or failures frequently occur. In addition, even if the operational environment is as ideal as assumed, it is not possible to grasp up-to-date and complete information about constituting components and devices to conduct optimal control of a large-scale and complex information system.

Bio-inspired information and communication technology, Bio-ICT in short, is our solution to the current and emerging problems of information and communication systems. Biological systems are inherently scalable, adaptive, and robust. In general, bio-inspired systems would be non-optimal, unstable, slow, and less manageable than optimally designed and rigorously controlled conventional systems. However, there are many additional advantages. Their overhead is low, that is, consumes less energy, computational power, and memory. Because of simplicity, they are less likely to suffer from program bugs. Sustainable information and communication systems do not pursue optimal performance in the assumed or controlled operational conditions. Instead, they keep operating under dynamically changing and unexpected environments.

For example, a swarm of ants goes to a food source and returns to a nest in a line by selecting a shorter path at a branch. However, there is no centralized control or global view of the area. Instead, each ant decides behavior, i.e. moving direction, depending on simple rules and local information that it can evaluate and observe. As a consequence of interaction among ants, they autonomously make a short trail. Such phenomena, where a globally coordinated pattern, structure, or behavior emerges from mutual interaction among simple and autonomous entities, is called “self-organization”. Self-organization is fully distributed and known to be scalable, adaptive, and robust. By adopting self-organization principles, we can establish such an information and communication system which accomplishes not necessarily optimal but near-optimal or sufficient performance without centralized optimal control even under the influence of simultaneous failures. Another example of promising biological principles is the brain. A tremendous number of simple and tiny devices, called neurons, constitute this highly complex and amazing organ. It consumes only 10 to 30 W but performs more robust and complex computation than an energy-consuming high-performance computer. By adopting brain-inspired models, we can expect highly energy-efficient and robust communication and information processing.

We do not only mimic the superficial behavior or structure of biological systems. Instead, we analyze the principles behind their emerging behavior and characteristics and apply them to ICT. More specifically, we build, evaluate, and adopt mathematical models explaining biological behavior. Through such an approach, we can correctly understand and interpret essential features of biological systems and then appropriately and effectively establish bio-inspired systems. For this purpose, we cooperate with biologists, neuroscientists, mathematicians, physicists, and engineers of academic, research, and industrial institutes.