The emergency medical assistant system for future human interplanetary traveling such like explore Mars or other deep space traveling requires substantial rethinking. Currently, on the International Space Station (ISS), the crew is provided with a paper manual to give guidance in the event of a medical need. For any condition, the physicians at the Mission Control Center (MCC) use voice communication refer the astronaut to a certain page of the manual and let the astronaut following the procedures list on the manual. But a printed manual is bulky and has a lot of limitations.

     In the future deep space traveling, the time period of human stay in space station will range from months to years. The distance between space station and earth may be hundreds of million miles. The communication between space station and earth will have up to 30 minutes round delay and predicated short period time of “black out” (communication link broken because of planet movement). The medical equipment installed on the spacecraft may be different at different mission. All those characters decide the Emergency Medical Assistant System (EMA) for future Space Station is different from the “space emergency telemedicine” that NASA relied on which generally involve the physician on the ground talking with the astronaut on the space station. This EMA system should be a kind of distributed decision support system, with some decision processing functions residing in space and others on earth. 

In this research, we proposed a multi-agent architecture for developing distributed artificial intelligent medical support system. All the agents in the system can intelligently, autonomously and collaboratively work to monitor the life critical data of the stuff on board, make an optimized emergency decision (treatment) according the status of environment in space station (such as the communication status, the available medical equipment, and so on), collect necessary information, send back to ground across delay communication link and issue last treatment which approved by human medical team on ground. Thereby avoiding fatal delay for getting treatment from ground during emergency situation. To demonstrate the architecture of this Emergency Medical Assistant System, a demo multi-agent system was successfully developed.