The many research robots developed generally classified into two categories:
1. Humanoid robot research platform is generally focuses on bipedal locomotion or human-robot interaction before manipulation, resulting in robots that are not designed to demonstrate practical manipulation performance.
2. Robot arm-and-gripper test beds that allow research on manipulating real objects. These research prototypes or modified industrial robots are not consumer-grade, human-safe or robust enough to be considered true development platforms.
Successful robot designs result from a tight coupling with the specific application space targeted by the developer. Robots designed, for example, for industry, space, surgery, and rehabilitation have significantly different operational criteria. The application space driving the design of this robot is broadly defined as human scale manipulation tasks in human environments.
Achieving functionality and safety in the variety of unstructured environments encountered in this application space presents a challenge since system behavior is dictated by software, not a human operator.
Software does not exist yet to perform the decision making humans do when dealing with safety, though this is an active area of research. To advance the area of personal robotics, for which assuring human safety is not optional and can not be handled through software alone, the mechanical design of the robot must be the ultimate safeguard.
No comments:
Post a Comment