Until quite recently, MPC was mainly used for chemical process control in industries like oil refining and petrochemicals, where many variables are in play, but system response is slow and the scope of controller computation is limited. However, MPC-based control technologies are now finding applications in other sectors, such as robot control, urban air mobility (UAM), self-driving cars, aerospace, integrated building energy control, smart grids, and financial engineering. Dramatic advances in sensing technologies, such as GPS, cameras, radar, and LiDAR, have increased the amount of high-quality data that can be used to predict the operating environment, and thanks to improved CPU performance, large amounts of data can be rapidly processed.
Yet MPC has rarely been applied in the automotive industry because of the challenges involved in computing a wide range of control variables, such as driving environment and conditions, in real time. As a result, model predictive control was considered less effective for automobiles. However, in recent years, with advances in the engine management system, or EMS, which analyzes data such as fuel quantity, air quantity, and exhaust pressure, accurate gaging of the current state is within reach. Moreover, as the accuracy of the MPC model has improved, applications for MPC-based technology are increasing.