Abstract: The current neuromotor rehabilitation setup is insufficient in the sense that the therapy depends heavily on therapists, the information associated with the patient's movement is underused, and the patient's assessment is primarily subjective. The neurorehabilitation research demands the rehabilitation engineers to advance the therapeutic tools that could improve both clinical effectiveness and cost effectiveness. With the development of exciting knowledge and techniques, the neuromotor rehabilitation setup can be changed into an assistive, interactive, and automatic rehabilitation environment in which the quantified information flows, constructing a network between machine, computer, and human. This study focuses on the development of a novel rehabilitation device to facilitate the neurologically injured patients to train the upper-extremity with functional tasks repetitively. In designed devices, the assistive therapeutic robot aids motor-impaired arm mechanically to perform repetitive task training. Meanwhile, the biofeedback, with an interactive and motivating computer-human interface, gives patients efficient augmented feedbacks through multimodal sensory cues and therefore enhances their active involvement in the training. The precisely quantified movement information is applied to control the robots automatically, displayed on the biofeedback computer-human interface, fed back to the patient to elicit his/her awareness on movement state, and presented to the therapist to evaluate the patient. Hence, this machine-computer-human inerative environment provides neurologically injured patients repetitive opportunities tot rain the task, motivates patients' volitional effort, enhances the sensorimotor integration with multimodal feedback and teaches patients task performance by interactive approach that are key ingredients to encourage the neural plasticity and motor function recovery. In addition, the information and automation technologies allow patients to receive task training even with less presence of therapists, improving the cost effectiveness of the therapy. The concepts and the designed therapeutic device demonstrated in this study are the first step to impel the realization of automatic rehabilitation environment or telemedical rehabilitation for home-use that can improve efficiency of neuromotor rehabilitation and the life of disabled people.