10 May 2021

Automated Terrain Classification for Vehicle Mobility in Off-Road Conditions

ABSTRACT:  The U.S. Army is increasingly interested in autonomous vehicle operations, including off-road autonomous ground maneuver. Unlike on-road, off-road terrain can vary drastically, especially with the effects of seasonality. As such, vehicles operating in off-road environments need to be informed about the changing terrain prior to departure or en route for successful maneuver to the mission end point. The purpose of this report is to assess machine learning algorithms used on various remotely sensed datasets to see which combinations are useful for identifying different terrain. The study collected data from several types of winter conditions by using both active and passive, satellite and vehicle-based sensor platforms and both supervised and unsupervised machine learning algorithms. To classify specific terrain types, supervised algorithms must be used in tandem with large training datasets, which are time consuming to create. However, unsupervised segmentation algorithms can be used to help label the training data. More work is required gathering training data to include a wider variety of terrain types. While classification is a good first step, more detailed information about the terrain properties will be needed for off-road autonomy.

1 Dec 2020

Snow-Covered Obstacles’ Effect on Vehicle Mobility

ABSTRACT:  The Mobility in Complex Environments project used unmanned aerial systems (UAS) to identify obstacles and to provide path planning in forward operational locations. The UAS were equipped with remote-sensing devices, such as photogrammetry and lidar, to identify obstacles. The path-planning algorithms incorporated the detected obstacles to then identify the fastest and safest vehicle routes. Future algorithms should incorporate vehicle characteristics as each type of vehicle will perform differently over a given obstacle, resulting in distinctive optimal paths. This study explored the effect of snow-covered obstacles on dynamic vehicle response. Vehicle tests used an instrumented HMMWV (high mobility multipurpose wheeled vehicle) driven over obstacles with and without snow cover. Tests showed a 45% reduction in normal force variation and a 43% reduction in body acceleration associated with a 14.5 cm snow cover. To predict vehicle body acceleration and normal force response, we developed two quarter-car models: rigid terrain and deformable snow terrain quarter-car models. The simple quarter models provided reasonable agreement with the vehicle test data. We also used the models to analyze the effects of vehicle parameters, such as ground pressure, to understand the effect of snow cover on vehicle response.

1 Sep 2020

PUBLICATION NOTICE: Seasonal Effects on Vehicle Mobility: High-Latitude Case Study

Abstract: Seasonality plays a key role in altering the terrain of many military operating environments. Since seasonality has such a large impact on the terrain, it needs to be properly accounted for in vehicle dynamics models. This work outlines a variety of static and dynamic seasonal terrain conditions and their impacts on vehicle mobility in an austere region of Europe. Overall the vehicles performed the best in the dry season condition. The thaw season condition had the most drastic impact on mobility with all but the heavy tracked vehicle being almost completely NOGO in the region. Overall, the heavy tracked vehicle had the best performance in all terrain conditions. These results highlight the importance of incorporating seasonal impacts on terrain into NRMM or any vehicle dynamics model. Future work will focus on collecting more data to improve the empirical relationships between vehicles and seasonal terrain conditions, thereby allowing for more accurate speed predictions.

29 Apr 2020

PUBLICATION NOTICE: Preliminary Assessment of Landform Soil Strength on Glaciated Terrain in New Hampshire

Abstract: Accurate terrain characterization is important for predicting off-road vehicle mobility. Soil strength is a significant terrain characteristic affecting vehicle mobility. Collecting soil strength measurements is laborious, making in-situ observations sparse. Research has focused on providing soil strength estimates using remote sensing techniques that can provide large spatial and temporal estimates, but the results are often inaccurate. Past attempts have quantified the soil properties of arid environments using landform assessments; yet many military operating environments occupy high latitude regions with landscapes dominated by glacial deposits. This study took preliminary strength measurements for glacial landforms deposited from the Laurentide Ice Sheet in New England. A range of common glacial landforms were sampled to assess shear strength, bearing capacity, and volumetric moisture content. Glacial outwash landforms had the highest average shear strengths, glacial deltas the lowest. There was a significant negative correlation between silt content and shear strength of the soil, a significant positive correlation between bearing capacity and clay content, and a significant negative correlation with sand content. Moisture content of soils was inversely correlated to the abundance of gravel in the deposit. This work provides initial insight to this approach on glaciated terrain, but continued sampling will provide more robust correlations.