Abstract

In this paper, we address the problem of automatically detecting and adapting to sensor failures, which is an important step towards building long-lasting survivable software. We present a novel constraint-based learning framework that performs joint sensor failure detection and adaptation. Our framework learns sensor relationships from historical data and expresses them as a set of constraints. These constraints then provide a joint view for detection and adaptation: detection checks which constraints are violated, and adaptation reconstructs failed sensor values. Additionally, we show that our framework can not only identify the mode of sensor failure but can also estimate the quality of the proposed adaptation. Our empirical studies on sensor data from the weather and appliance energy domains demonstrate the advantages of our approach over other methods.

Date

November 5, 2018

Authors

Yuan Shi, TK Satish Kumar, Craig A Knoblock

Conference

2018 IEEE 30th International Conference on Tools with Artificial Intelligence (ICTAI)

Pages

328-335

Publisher

IEEE