The use of military vehicles during training results in soil disturbance and vegetation loss, with subsequent increases in soil erosion rates, sedimentation in streams, habitat degradation, and numerous other secondary effects. The National Environmental Policy Act (NEPA) requires federal agencies to evaluate the implications of their plans, policies, programs, and projects. However, accurate assessment of military training impacts is limited by the technical data available to support the assessments. This project demonstrated the use of the Vehicle Dynamics Monitoring and Tracking System (VDMTS) to assess and predict military vehicle maneuver training impacts for use in land management decision making and NEPA documentation. The objective of this project was to demonstrate and validate VDMTS and its components through a series of controlled field studies and live tracking events. A controlled field study was used to demonstrate and validate that the hardware can sufficiently characterize vehicle dynamic properties (turning radius and velocity) to accurately predict site impacts (area impacted, vegetation loss, and rut depth [RD]). A controlled field study was used to demonstrate and validate the accuracy of VDMTS impact models in predicting area impacted, vegetation loss, and RD for a range of vehicles. Field studies tracking live training exercises and subsequent field measurements were used to demonstrate and validate the VDMTS hardware and model performance in predicting site impacts.

Technology Description

The VDMTS approach is composed of three components: (1) vehicle impact models, (2) vehicle tracking hardware and software, and (3) vehicle tracking data analysis. The approach spatially characterizes short-term, direct impacts by monitoring individual vehicle locations and operating characteristics. These dynamic characteristics are used to predict area impacted, vegetation loss, and RD based on vehicle type and location. Results are summarized to characterize training land use patterns and quantify the severity of the training impacts.

The process-based impact models predict terrain impacts caused by wheeled and tracked vehicles in terms of percent vegetation cover loss (impact severity [IS]), disturbed width (DW), and RD. Impact models predict site impacts based on vehicle weight and type, vehicle dynamic properties, and soil properties (soil strength). The process uses vehicle tracking systems to determine vehicle location and dynamic operating characteristics (i.e., turning radius and velocity). The VDMTS hardware consists of a Global Positioning System (GPS) receiver integrated with low-cost inertial sensors. These sensors enable measurement of vehicle kinematics, dynamics, and other parameters of interest that enable accurate modeling of environmental impact. The system thereby provides vehicle dynamics data and positional information at all times, even when GPS is unavailable.

The vehicle tracking data are analyzed and summarized into formats appropriate for land management decisions. Analysis routines include: (1) identification of individual and unit tracking patterns, (2) identification of on- and off-road use patterns, (3) identification of existing and emerging trail networks, (4) vegetation loss estimates, (5) identification and prioritization of Land Repair and Maintenance (LRAM) sites, and (6) development of data for carrying capacity models.

Demonstration Results

This project tested and validated each aspect of the VDTMS process at multiple levels, specifically, accuracy of the hardware and models in combination, durability of the hardware under multiple training events, ease of use of the VDTMS process, and ability to make land-use decisions based on the VDMTS collected and summarized data. The following quantitative metrics were tested to assess each aspect of VDMTS performance: (1) accurate VDMTS hardware measurement of vehicle dynamic properties, (2) accurate VDMTS impact model predictions of site impacts under controlled conditions, (3) accurate VDMTS hardware measurement of vehicle static and dynamic properties, (4) accurate VDMTS model predictions of site impacts during live training, (5) VDMTS hardware durability (in single live training event), (6) VDMTS hardware durability over 14 live training events, (7) ease of system use, and (8) quality and accuracy of data for land-use decisions.

Of the hardware performance metrics cited above, 1, 3, and 5-8 were met. Metrics 2 and 4, accurate VDTMS impact model predictions in controlled and live events, did not meet the success criteria initially proposed. The demonstrated average error for DW was 14.9 cm, and the average error for vegetation removal was -1.8%. These results are comparable with existing site and vehicle-specific empirical model predictions, thus reducing the need to develop models for each site. This validates the use of the theoretical models for impact prediction.

The system met most of the metrics established. While it failed to meet some metrics, it still performed as well as previous methods in characterizing vehicle impacts, reducing the relative cost and time required. Project success was indicated by the use of data obtained from the system by the installation hosts as well as their quick implementation of the technology. Through the course of the project, installation Integrated Training Area Management (ITAM), Environmental, Directorate of Public Works (DPW), and training groups used results from this study. Data collected were used in land management and vehicle mobility and power models. Study results also informed training and regulating decisions.

Implementation Issues

Implementation of the VDMTS approach will generally be driven by installation land manager requirements. Despite installation acceptance of the process, a lack of continuity from turnover in the ITAM and Environmental installation branches may result in issues incorporating the system into those programs. An additional option for implementation involves using existing military standard systems (the Army’s Blue Force Tracking [BFT] and the National Guard’s Deployable Force-on-Force Instrumented Range System [DFIRST]), which obtain vehicle location and time data on live training events for post-event analysis. This option would decrease costs by reducing the time and equipment required to obtain vehicle tracking data. However, this process is still being developed and is not available widely at this time. The VDMTS technology has been and will continue to be valuable in obtaining data to estimate impacts from military training.