A Fault Tolerant Intelligent Robotic Assembly Framework (University of HK)

In conventional automated assembly, a robot is used to pick up parts and assemble them under some pre-taught paths. Yet for tight-tolerance assemblies, slight position errors and disturbances will often hamper the reliability and quality of the operation. A possible advancement to such problem is to enhance the intelligence of the robot by introducing the sense of the environment in which the robot is operating, and to incorporate the power of evaluating the situation such that the best manoeuvre can be planned. The project is aimed at enhancing the dexterity and the ability to recover from fault conditions of robots used in automatic assembly operations. Resulting force and torque information obtained while assembly parts make contacts with their environment. In particular, the research is focused onto the development of a framework for task-level intelligent assembly using an industrial robot to perform successful assembly tasks such as mating of work-pieces, with the ability to self-correct faults including misalignments and control errors. An Intelligent Robot Er r or Diagnosis