IMU (Inertial Measurement Unit) sensors were initially used in space and military contexts. However, with the advancement of technology, new generations of these sensors have emerged and are now finding applications in several fields. The development of Artificial Intelligence (AI) has been an essential part of this progress, driving improvements in the ability of IMUs to adapt to complex and variable environments, mainly thanks to the integration of neural networks. This synergy between IMU sensors and AI has allowed their use in applications such as infrastructure monitoring, covering the structural health of buildings, street mapping, and detecting irregularities in road infrastructure. In these areas, methods such as the use of Support Vector Machines (SVM) or classification and regression in Structural Health Monitoring (SHM) systems and real-time structural damage detectors using one-dimensional Convolution Neural Networks (CNN), and Artificial Neural Networks (ANN), among other applications, stand out. Finally, a case study is presented in which a CNN is used to classify speed bumps from acceleration data obtained by an IMU sensor. The CNN inputs consist of windows generated from the inertial sensor readings. The system proved to be efficient for real-time speed bump classification, achieving 91% accuracy in training tests. During real-time experimentation, the network achieves 90% accuracy when applying a probability threshold of 0.9 for the “speed bump” class.