Imitation Learning Employee Training for Artificial Intelligence ICC Service Robots

초록

The service robot industry is rapidly expanding to various fields such as healthcare, care, and logistics due to the aging, shortage of manpower, and spread of non-face-to-face services. However, due to the high unpredictability of the service environment, there is a limit to implementing stable and flexible robot behavior only with existing rule-based control. This curriculum is designed as a core technology for service robots to adapt to the environment and learn behavior through reinforcement learning of artificial intelligence robots. This technology has high theoretical difficulty and complexity in implementation, and there are many practical barriers such as simulation-real problem, data acquisition, and safety verification. In addition, there is a lack of professional manpower who can apply the technology to the corporate field, resulting in trial and error and decision-making risk. Therefore, it is necessary to provide systematic robot control education and practical experience to the incumbent based on reinforcement learning and video learning. The curriculum for imitation learning during service focused on solving problems that repeat the sex and improvement of reinforcement learning. Learners secure the stability and reproducibility of policies through stable log analysis, average compensation evaluation, and repetitive seed verification. As a result of the study, the core application fields of service robots were presented, the excellence of imitation learning-based technology and the applicability of education were confirmed, and the success rate of field application was analyzed at about 70%.

키워드

Artificial Intelligence ICCService RobotImitation LearningTraining of the IncumbentReinforcement Learning
제목
Imitation Learning Employee Training for Artificial Intelligence ICC Service Robots
저자
Moonhee ChoiMinsu Kim
DOI
10.7236/IJIBC.2026.18.3.182
발행일
2026-06
유형
Y
저널명
The International Journal of Internet, Broadcasting and Communication
18
3
페이지
182 ~ 193