Fault detection, isolation and recovery is an important aspect of spacecraft operations especially in case of unexpected faults and failures. Traditional fault diagnosis techniques on-board spacecraft are very conservative and do neither support context sensitive on-board decision-making nor goal-oriented behaviour. This paper argues that the rapidly increasing processing power enables the use of innovative methods although they might need more resources than were available in the past. As a promising method, the cognitive automation approach from the unmanned aerial vehicle domain is identified and a respective case study to investigate the suitability in spacecraft applications is proposed.
«Fault detection, isolation and recovery is an important aspect of spacecraft operations especially in case of unexpected faults and failures. Traditional fault diagnosis techniques on-board spacecraft are very conservative and do neither support context sensitive on-board decision-making nor goal-oriented behaviour. This paper argues that the rapidly increasing processing power enables the use of innovative methods although they might need more resources than were available in the past. As a pro...
»