Phillip J. Durst

A Real-Time, Interactive Simulation Environment for Unmanned Ground Vehicles: The Autonomous Navigation Virtual Environment Laboratory (ANVEL)

Modeling and simulation tools have become an integral part of modern engineering processes. In particular, accurate and efficient simulation tools are critical for the design, development, and testing of autonomous unmanned ground vehicles (UGVs). However, because of the complexity of the problem, many UGV simulators are computationally intensive, require expensive hardware to run, and are often not interactive or real-time. Those simulation environments that do provide users with some degree of interactivity and real-time or faster performance gain these features at the sacrifice of simulation fidelity, and these products often provide inadequate results. A new simulation environment for UGV design and development, called the Autonomous Navigation Virtual Environment Laboratory (ANVEL), has been created to address the need for a real-time, interactive, physics-based simulation environment for UGVs. ANVEL is able to meet this need by fusing readily available, off-the-shelf video game technology with high-fidelity, physics-based models. This paper presents the methodology used in developing ANVEL, an example use of ANVEL for development and testing of an autonomous UGV, and plans for the future development.

High fidelity sensor simulations for the virtual autonomous navigation environment

The Virtual Autonomous Navigation Environment (VANE) is a high-fidelity simulation environment for ground robotics. Physicsbased realism is the primary goal of the VANE. The VANE simulation incorporates realistic lighting, vehicle-terrain interaction, environmental attributions, and sensors. The sensor models, including camera, laser ranging, and GPS, are the focus of this work. These sensor models were designed to incorporate both internal (electronic) and external (environment) noise in order to produce a sensor output that closely matches that produced in real-world environments. This sensor output will allow roboticists to use simulation further into the development and debugging process before exposing robots to field conditions.

A Framework for Predicting the Mission-Specific Performance of Autonomous Unmanned Systems

While many methodologies have been proposed for calculating a quantitative level of autonomy for intelligent Unmanned Systems (UMS), no one definitive measure of autonomy or autonomous performance has been validated and adopted by the UMS community. Particularly for military applications, a simple performance metric that is based on the UMSs mission profile and is comparable between UMS systems is critical. This metric would not only help define the features a UMS needs to successfully perform its mission, both in terms of hardware and software, but also enable the use of UMS for a broader range of applications at an increased level of autonomy. This paper presents the development of a new methodology for calculating a single-number performance metric for autonomous UMS, and this metric is called the Mission Performance Potential (MPP). Rather than a retroactive measure of UMS performance and autonomy level for one iteration of a given scenario, the MPP separates autonomy level and mission performance to provide a predictive measure of a UMSs expected performance for a mission set and level of autonomy. As an example application, the MPP is calculated for an Unmanned Aerial Vehicle (UAV) performing a target tracking mission, and this MPP value is compared to the results of field-testing with this system.

The Need for High-Fidelity Robotics Sensor Models

Simulations provide a safe, controlled setting for testing and are therefore ideal for rapidly developing and testing autonomous mobile robot behaviors. However, algorithms for mobile robots are notorious for transitioning poorly from simulations to fielded platforms. The difficulty can in part be attributed to the use of simplistic sensor models that do not recreate important phenomena that affect autonomous navigation. The differences between the output of simple sensor models and true sensors are highlighted using results from a field test exercise with the National Robotics Engineering Centers Crusher vehicle. The Crusher was manually driven through an area consisting of a mix of small vegetation, rocks, and hay bales. LIDAR sensor data was collected along the path traveled and used to construct a model of the area. LIDAR data were simulated using a simple point-intersection model for a second, independent path. Cost maps were generated by the Crusher autonomy system using both the real-world and simulated sensor data. The comparison of these cost maps shows consistencies on most solid, large geometry surfaces such as the ground, but discrepancies around vegetation indicate that higher fidelity models are required to truly capture the complex interactions of the sensors with complex objects.

A general model for inferring terrain surface roughness as a root–mean–square to predict vehicle off–road ride quality

Vehicle maximum speed for off-road operations is limited by the absorbed power via vertical acceleration to the driver for a given terrain Root-Mean-Square surface roughness (RMS). RMS calculation requires centimetre-scale terrain elevation data; however, previous work by the authors has shown that RMS can be modelled using a 5-metre terrain profile’s Fractal Dimension (FD) and Power Spectral Density (PSD) DC offset. Presented is a study of the effects of surface elevation data resolution on the model. Forty-nine ride courses were down-sampled from 30 centimetre to 0.91, 1.83, 2.74, 3.66, 4.57, 5.49, 6.40, 7.32, and 8.23 metre spacings, and an RMS model at each spacing was generated using linear regression techniques. The effects of data resolution on the RMS model were studied, and a continuous model for RMS as a function of FD and DC offset across elevation data resolutions for up to 7 metre sample spacing was developed. Results of the model’s use in predicting off road military vehicle mobility are presented.

A probabilistic model for simulating the effect of airborne dust on ground-based LIDAR

Field and laboratory measurements of Light Detection and Ranging (LIDAR) sensor interactions with dust have been performed for two types of common ground-based LIDAR sensors. A strong correlation (R2 > 0.95) between the probability for a return from the dust and the optical depth of the dust was found in the analysis. Based on the experimental correlation, a probabilistic model for LIDAR interactions with dust was developed and verified in field experiments. Finally, the model was integrated into a high-fidelity ray-tracing simulation of LIDAR systems.

A historical review of the development of verification and validation theories for simulation models

Modeling and simulation (M&S) play a critical role in both engineering and basic research processes. Computer-based models have existed since the 1950s, and those early models have given way to the more complex computational and physics-based simulations used today. As such, a great deal of research has been done to establish what level of trust should be given to simulation outputs and how to verify and validate the models used in these simulations. This paper presents an overview of the theoretical work done to date defining formal definitions for, and methods of, verification and validation (V&V) of computer models. Simulation models are broken down into three broad categories: analytical and simulation models, computational and physics-based models, and simulations of autonomous systems, and the unique theories and methods developed to address V&V of these models are presented. This paper also presents the current problems in the theoretical field of V&V for models as simulations move from single system models and simulations to more complex simulation tools. In particular, this paper highlights the lack of agreed-upon methods for V&V of simulations of autonomous systems, such as an autonomous unmanned vehicles, and proposes some next steps needed to address this problem.

Simulation of a multispectral, multicamera, off-road autonomous vehicle perception system with Virtual Autonomous Navigation Environment (VANE)

We present a case-study in using specialized, physics-based software for high-fidelity environment and electro-optical sensor modeling in order to produce simulated sensor data that can be used to train a multi-spectral perception system for unmanned ground vehicle navigation. This case-study used the Virtual Autonomous Navigation Environment (VANE) to simulate filtered, multi-spectral imaging sensors. The VANE utilizes ray-tracing and hyperspectral material properties to capture the sensor-environment interaction. In this study we focus on a digital scene of the ERDC test track in Vicksburg, MS that has extremely detailed representation of the vegetation and ground texture. The scene model is used to generate imagery that simulates the output of specialized terrain perception hardware developed by Southwest Research Institute, which consists of stereo pair of 3-channel cameras. The perception system utilizes stereo processing, the multi-spectral responses, and image texture features in order to create a 3-dimensional world model suitable for offroad vehicle navigation, providing depth information and an estimated terrain class label for every pixel by utilizing machine learning. While the process of training the perception system generally involves hand-labeling data collected through manned missions, the ability to generate data for certain environments and lighting conditions represents an enabling technology for deployment in new theaters. We demonstrate an initial capability to simulate data and train the perception system and present the results compared to the system trained with real-world data from the same location.

Development of a non-contextual model for determining the autonomy level of intelligent unmanned systems

A simple, quantitative measure for encapsulating the autonomous capabilities of unmanned systems (UMS) has yet to be established. Current models for measuring a UMS’s autonomy level require extensive, operational level testing, and provide a means for assessing the autonomy level for a specific mission/task and operational environment. A more elegant technique for quantifying autonomy using component level testing of the robot platform alone, outside of mission and environment contexts, is desirable. Using a high level framework for UMS architectures, such a model for determining a level of autonomy has been developed. The model uses a combination of developmental and component level testing for each aspect of the UMS architecture to define a non-contextual autonomous potential (NCAP). The NCAP provides an autonomy level, ranging from fully non- autonomous to fully autonomous, in the form of a single numeric parameter describing the UMS’s performance capabilities when operating at that level of autonomy.

A history and overview of mobility modeling for autonomous unmanned ground vehicles

Autonomous unmanned ground vehicles (UGVs) are beginning to play a more critical role in military operations. As the size of the fighting forces continues to draw down, the U.S. and coalition partner Armed Forces will become increasingly reliant on UGVs to perform mission-critical roles. These roles range from squad-level manned-unmanned teaming to large-scale autonomous convoy operations. However, as more UGVs with increasing levels of autonomy are entering the field, tools for accurately predicting these UGVs performance and capabilities are lacking. In particular, the mobility of autonomous UGVs is a largely unsolved problem. While legacy tools for predicting ground vehicle mobility are available for both assessing performance and planning operations, in particular the NATO Reference Mobility Model, no such toolset exists for autonomous UGVs. Once autonomy comes into play, ground vehicle mechanical-mobility is no longer enough to characterize vehicle mobility performance. Not only will vehicle-terrain interactions and driver concerns impact mobility, but sensor-environment interactions will also affect mobility. UGV mobility will depend in a large part on the sensor data available to drive the UGVs autonomy algorithms. A limited amount of research has been focused on the concept of perception-based mobility to date. To that end, the presented work will provide a review of the tools and methods developed thus far for modeling, simulating, and assessing autonomous mobility for UGVs. This review will highlight both the modifications being made to current mobility modeling tools and new tools in development specifically for autonomous mobility modeling. In light of this review, areas of current need will also be highlighted, and recommended steps forward will be proposed.