Space Intelligence Evolution with Autonomic Preference Learning

space intelligenceSpace intelligence-based service providing in a hyper-connected society by autonomous IoT objects

  • As our society is hyper-connected, billions of distributed smart objects are connected to each other and share their data with anything, anytime, anywhere in IoT environments. To enable the hyper-connection, existing IoT frameworks collect IoT big data to a cloud server and statistically analyze the embodied characteristics. However, this approach may cause an information loss, a system overhead, or eventually, the cloud server may become a single-point-of-failure. To prevent the information loss and guarantee system scalability, it is necessary to separate IoT data and services by a smaller but meaningful unit. The ‘meaningful unit’ here indicates a unit in which user intentions rise and surrounding smart objects collaborate to support them. In this project, we define a ‘place’ as the meaningful unit which keeps the IoT data and services in a manageable size. To realize this goal, we need to develop a technology that distributed intelligent objects determine situations by themselves based on the knowledge accumulated in the space and autonomously collaborate with other objects.

Why do we need to recognize user intentions and optimize smart object collaborations through space intelligence?

  • An IoT framework should provide each smart object services according to various users’ intentions seamlessly. The scale urban IoTs became too large for users to spontaneously detect and control nearby smart objects. In addition, a more advanced level of space intelligence is required to perform an adaptive service providing like: the composition of intelligent objects included in the space changes, the existing service providing method also be converted. Therefore, automated / streamlined /optimized technologies that provide user-preference services by cooperating adaptive intelligence objects with each other in real time based on a knowledge which accumulated in the space in large scale are needed.

Core Research Topics to Realize Space Intelligence:

  • An Ontology-based Knowledge Extension for Better Space Situation Awareness
  • Self Evolution of Space Knowledge based on Place Big Data
  • Identifying and Tracking User Intentions based on Data-driven Persona Generation
  • An Autonomous Reconfiguration of IoT Collaboration Services according to User Intentions and IoT State Changes
  • A Context-aware Intelligent Messaging for Optimizing IoT Collaboration
  • Modeling and Optimizing an Adaptive Collaboration of Distributed IoTs according to User Preference
  • Development of Various Service Scenarios and Real IoT-enabled Testbeds