Characterization of Neural Networks Automatically Mapped on Automotive-grade Microcontrollers

TL;DR

This paper evaluates the implementation of neural networks on automotive-grade microcontrollers, demonstrating their effectiveness in vehicle security and battery management through two case studies.

Contribution

It introduces a framework for deploying neural networks on automotive microcontrollers, enabling edge computing for automotive applications.

Findings

Neural networks can be effectively mapped on automotive microcontrollers.

The framework achieves efficient intrusion detection and battery capacity estimation.

Potential for real-time, low-power automotive AI applications.

Abstract

Nowadays, Neural Networks represent a major expectation for the realization of powerful Deep Learning algorithms, which can determine several physical systems' behaviors and operations. Computational resources required for model, training, and running are large, especially when related to the amount of data that Neural Networks typically need to generalize. The latest TinyML technologies allow integrating pre-trained models on embedded systems, allowing making computing at the edge faster, cheaper, and safer. Although these technologies originated in the consumer and industrial worlds, many sectors can greatly benefit from them, such as the automotive industry. In this paper, we present a framework for implementing Neural Network-based models on a family of automotive Microcontrollers, showing their efficiency in two case studies applied to vehicles: intrusion detection on the Controller Area Network bus and residual capacity estimation in Lithium-Ion batteries, widely used in Electric Vehicles.