Neuropsychology is the study of how the brain relates to behavior, emotion, and cognition. Clinical neuropsychologists evaluate the behavioral effects of neurological and developmental disorders stemming from brain injury, strokes, multiple sclerosis, Alzheimer’s disease, and Parkinson’s disease. Millions of Americans are currently living with these cognitive disorders, including a growing number of veterans returning from Iraq with brain injuries. The disorders often result in cognitive impairments that make it difficult to plan daily activities and stay on task, affecting independence, quality of life, and employment.
In the early 1990s, Richard Levinson, a NASA contractor and senior researcher in Ames Research Center’s Artificial Intelligence Research Branch and Autonomous Systems Group, took the science in an entirely different direction when he folded it into his NASA work.
Levinson, who had previously received a Space Act Award for contributing to the development of a prototype autonomous thermal control system for the International Space Station, initially learned about neuropsychology accidentally. Moving into a new apartment in 1986, he received a neuropsychology course program in the mail that was actually intended for the previous occupant. The topic intrigued Levinson, and the pursuit to learn more was on.
As Levinson learned about emerging neuropsychological models of human planning, he continued researching computer models of automatic planning. A central concern for both fields is that plans often change when surprises occur.
Neuropyschologists study how this integrated planning and execution breaks down as the result of cognitive impairment, but they do not know exactly how planning occurs in the brain. On the other side, computer scientists can build a planning system, but have a limited understanding of how to integrate planning with execution monitoring and error recovery.
Levinson studied the neuropsychology of human planning and applied this knowledge to his NASA research in order to increase autonomy for spacecraft and robots. Since spacecraft and robots operate in uncertain conditions, they cannot be preprogrammed for every activity, so there are times that they must be responsible for their own “health” and safety. Further, as NASA’s missions grow more complex, so does the Agency’s need for machines that can exhibit a higher degree of independence and execute improvised actions in novel situations where preprogrammed commands will not work.
In 1995, Levinson published peer-reviewed research papers in computer science and neuropsychology journals, describing an artificially intelligent planning and reaction model founded on neuropsychological theories of human behavior. This planning and reaction model was based on the functioning of the human brain’s frontal lobes, which play a part in memory, motor skills, planning, decision making, and socialization, among other functions.
Levinson has since received three patents for the technology, pertaining to activity planning and cueing methods with execution monitoring and error correction.
While Levinson and NASA continue to investigate this advanced computer model for future missions, the technology has already made its terrestrial debut in the form of a powerful cueing and scheduling aid to help people with a wide range of cognitive, attention, and developmental disorders.
Levinson received initial funding from NASA and his contracting company, Recom Technologies Inc., of Roseville, California, to research the commercial potential of his artificially intelligent planning reaction model to serve as a tool for helping individuals suffering from various forms and levels of brain impairment. In 1993, the chief of Ames’ Artificial Intelligence Research Branch suggested that Levinson contact Santa Clara Valley Medical Center, which hosts a nationally acclaimed rehabilitation and research center that specializes in brain injuries, to see if the hospital was interested in a research collaboration. Levinson heeded the advice and found a valuable partner in the medical center. This partnership led to further development of Levinson’s technology and funding to support clinical research from the U.S. Department of Education’s National Institute on Disability and Rehabilitation Research.
In 1996, Levinson founded Attention Control Systems Inc., in Mountain View, CA, to produce and market this NASA spinoff creation.
How it Works
Attention Control Systems now offers people with memory, attention, and cognitive disorders a computerized personal planning device to help them stay on task by overcoming limitations in planning and fulfilling their daily schedules. The device, called the Planning and Execution Assistant and Trainer, or PEAT, is a pocket-sized PDA, complete with a graphical display, touchscreen controls, an electronic calendar, an address book, and a built-in phone. The functionality of PEAT, however, transcends that of a regular PDA scheduling device. PEAT cues users to start or stop scheduled activities, monitors their progress, and adjusts schedules as necessary in response to delays or calendar changes. It uses the automatic planning model developed for NASA to make automatic adjustments to daily plans when a situation changes. Most PDA systems lack this flexibility, requiring their users to manually re-plan and update schedule data when changes occur.