What is AIR?

The AIR project began its exploration in 2017 to build artificial intelligence for human-care robots. The official title of the project is “Development of Human-care Robot Technology for Aging Society”. The project is funded by the Korean gevernment agency, called MSIP(Ministry of Science, ICT and Future Planning).

Project Objectives

The project aims to develop artificial intelligence technologies for socially assistive robots that can provide personalized services in the areas of healthcare, affective interaction, and cognitive/informative support. The target consumers considered are the elderly people who can maintain independent living but are generally weak and incompetent in some capabilities.

The technologies should be viable to continuously monitor long-term changes in personal preferences, health conditions, daily activity patterns and social relationships to better understand the elderly people in various aspects and make qualified reports to families or dedicated care givers.

Work Packages

R1: Robot AI for Personal Profiling

  • Biometrics and Soft-Biometrics Recognition for the Elderly
  • Daily Activity and Interaction Gesture Recognition
  • Personal Objects and Information Learning
  • Robot Camera-based Real-time Human Detection and Tracking in In-House Environment
  • Building Large-scale Real-world Datasets
  • Building Large-scale Virtual Datasets
    • Virtual Elderly Human Modeling
    • Virtual Elderly People’s Home Environment Modeling
    • Virtual Data Synthesis with Auto-Created Variations

R2: Interpersonal Intelligence for Robots

  • Learning Interpersonal Intelligence by Observing Human Interactions
  • Generation of Robot Actions Tailored to Elderly Users
  • Understanding Sentiments in Elderly Users’ Daily Activities
  • Video Understanding and Description Generation

R3: Health Abnormality Symptoms Detection

  • Multi Sensor-based Action Recognition and Lifelogging
  • Health Abnormality Detection by Action Novelty Analysis

R4: User Intention Understanding

  • Neuro-Symbolic Fusion-based Action Intention Understanding
  • Compound Knowledge-based ADL/IADL Recognition

R5: Robot AI Evaluation Models

  • Systematic Categorization of Robotic Intelligence
  • Intelligence Quality Evaluation Models for Robotic Intelligence

R6: Systems and Services

  • Human-care Robot System Integration
  • Service Management S/W Tools
  • Human-case Services Suite for Elderly Users (Emotional/Life/Health/Cognitive Care Services)
  • Building and Managing Testbeds and Living Labs


All the publications, software and datasets described in this site were supported by the ICT R&D program of MSIP/IITP. [2017-0-00162, Development of Human-care Robot Technology for Aging Society]


  • Minsu Jang (minsu(at)etri.re.kr)