Extending MAPE-K with Data Augmentation to Mitigate Data Scarcity

The deployment of Autonomous swarms in remote or hazardous environments, such as space exploration presents new challenges to data collection. The distance to earth necessitates full autonomous and autonomic operation without human direction. Autonomic Computing and the MAPE-K loop provide a control structure for self-management. An autonomic system processes internal and external data and uses it to base decisions on. In the event that the data is scarce or missing, this reduces the reliability of the system. This paper addresses the issue of data scarcity by presenting and evaluating the effectiveness of using numeric interpolation to fill in missing data produced by a multi-robot swarm. The interpolated data is used to complete an existing dataset that is then passed through a data generation and evaluation pipeline. The results show that interpolation and generation can produce high quality synthetic data that could be used to mitigate data scarcity.