Title :
Execution error recovery for planning systems
Author :
Chang, Kai Hsiung ; Edhala, Mammatha
Author_Institution :
Comput. Sci. & Eng., Auburn Univ., AL, USA
Abstract :
A system is presented that is capable of handling failures during execution. The error recovery process consists of error detection, error identification and plan repair. The problem of error identification has been solved by using the condition-oriented approach, which is the central theme of this study. Error classification is necessary for accurate modification of a failing plan. The classification process is accomplished by checking the role of a failed condition in the plan. If all the various conditions in the plan are stored during the plan-generation phase, the affected part of the plan can be easily located. Four error categories are used, and the modification processes are described
Keywords :
artificial intelligence; system recovery; artificial intelligence; condition-oriented error identification; error classification; execution error recovery; planning systems; Artificial intelligence; Computer errors; Computer science; Condition monitoring; Phase detection; Problem-solving; Process planning; Tree data structures;
Conference_Titel :
Computers and Communications, 1988. Conference Proceedings., Seventh Annual International Phoenix Conference on
Conference_Location :
Scottsdale, AZ
Print_ISBN :
0-8186-0830-7
DOI :
10.1109/PCCC.1988.10127