This paper introduces the main progress of research on anti-fall gait balance training robots at home and abroad, and provides reference ideas for the research of anti-fall gait balance training rehabilitation robots and other fields of medical-industrial integration. The latest progress in the research of lower limb assisted rehabilitation robot is mainly summarized from four aspects, including:robot configuration; mechanism control strategy and algorithm; gait data collection method and gait data analysis method. In the current research, the market positioning of robots is a challenge faced by the current academic community. At present, various mainstream solutions have different focuses and different user segments. However, the main target group, the elderly, has strong user differences, so users have high personalized training needs. In the future, it is necessary to further refine the positioning of the gait balance training robot to meet the personalized training needs of different users at different stages in different situations, which is expected to improve the overall level of gait balance training robots.