Mid Sweden University

Inertial Measurement Units (IMUs) are widespread in many different applications, such as automotive, drone, robotics, smartphones, and many others. MEMS-based IMUs (i.e., IMU-based on Micro-Electro-Mechanical Systems technology) are the best solution to achieve effective and efficient monitoring and diagnostic with low cost, low power, and low dimension. The data acquired by the MEMS sensors integrated into IMUs are usually managed by suitable filtering and positioning algorithms. These algorithms are used to correct the system's attitude and vehicle path in which the IMU is installed. They also provide redundant information used for online diagnosis. Unfortunately, testing procedures specifically developed to characterize both IMU and filtering algorithms under stress are not yet available. Focusing the attention on the temperature's influence, the characterization of IMUs, considering the real scenario in which the system is operating, is a fundamental topic that must be deeply studied and discussed during a diagnostic system design. For this reason, this paper proposes a test plan for the performance analysis of MEMS-based IMUs under temperature-based stress tests. Beyond the primary analysis of reliability and functional operating, the effects of temperature miss-compensation have been evaluated on two common filtering algorithms employed for positioning applications and working on data coming from commercial IMUs implemented in the automotive field. The experimental results highlight how temperature plays a fundamental role in the behaviors of the investigated positioning algorithms. © 2021 IEEE.