Search Results

The Powertrain Analysis and Computational Environment (PACE) is a forward-looking powertrain simulation tool that is ready for a High-Performance Computing (HPC) environment. The code, written in C++, is one actor in a comprehensive ground vehicle co-simulation architecture being developed by the CREATE-GV program. PACE provides an advanced behavioral modeling capability for the powertrain subsystem of a conventional or hybrid-electric vehicle that exploits the idea of reusable vehicle modeling that underpins the Autonomie modeling environment developed by the Argonne National Laboratory. PACE permits the user to define a powertrain in Autonomie, which requires a single desktop license for MATLAB/Simulink, and port it to a cluster computer where PACE runs with an open-source BSD-3 license so that it can be distributed to as many nodes as needed.

While convolutional neural networks (CNNs) have revolutionized ground-vehicle autonomy in the last decade, this class of algorithms requires large, truth-labeled data sets to be trained. The process of collecting and labeling training data is tedious, time-consuming, expensive, and error-prone. In order to automate this process, an automated method for training CNNs with simulated data was developed. This method utilizes physics-based simulation of sensors, along with automated truth labeling, to improve the speed and accuracy of training data acquisition for both camera and LIDAR sensors. This framework is enabled by the MSU Autonomous Vehicle Simulator (MAVS), a physics-based sensor simulator for ground vehicle robotics that includes high-fidelity simulations of LIDAR, cameras, and other sensors.

Research interests in autonomous driving have increased significantly in recent years. Several methods are being suggested for performance optimization of autonomous vehicles. However, weather conditions such as rain, snow, and fog may hinder the performance of autonomous algorithms. It is therefore of great importance to study how the performance/efficiency of the underlying scene understanding algorithms vary with such adverse scenarios. Semantic segmentation is one of the most widely used scene-understanding techniques applied to autonomous driving. In this work, we study the performance degradation of several semantic segmentation algorithms caused by rain for off-road driving scenes. Given the limited availability of datasets for real-world off-road driving scenarios that include rain, we utilize two types of synthetic datasets.