Prognostics and Health Management Data Handling by Critical Tasks on Multi-Core Platforms

The collection of Prognostics and Health Management (PHM) data is very important for predictive maintenance of equipment (such as used in airplanes, automobiles, industrial machines). The data is provided by a variety of sensors that monitor the equipment, but the same data is used by the controllers that control the equipment. The deterministic functioning of the controllers which are typically run on multicore platforms is critical for certification purposes. The problem is then that the collection of the PHM data should interfere the least with the controllers operation. In this work, we implement a testbed using a quad-core ARM-based multi-core platform and report experimental results on the impact of the PHM data handling on the communication delay and execution delay of the critical controller tasks. We also provide insights about the system behavior.