A Standard Test Method for Evaluating Navigation and Obstacle Avoidance Capabilities of AGVs and AMRs

Abstract

Adam Norton1, Peter Gavriel1, and Holly Yanco1 ABSTRACT Automatic guided vehicles (AGVs) and autonomous mobile robots (AMRs) are now ubiquitous in industrial manufacturing environments. These systems all must possess a similar set of core capabilities, including navigation, obstacle avoidance, and localization. However, there are few standard methods to evaluate the capabilities and limitations of these systems in a way that is comparable. In this paper, a standard test method is presented that can be used to evaluate these capabilities and can be easily scaled and augmented according to the characteristics of the system under test. The test method can be configured in a variety of ways to exercise different capabilities, all using a common test apparatus to ease test set up and increase versatility. For each test configuration, conditions are specified with respect to the a priori knowledge provided to the system (e.g., boundary and/or obstacle locations) and the obstacles in the environment. Robustness of system capabilities is evaluated by purposefully introducing misalignment between the characteristics of the physical and virtual environment (e.g., providing representations of obstacles in the system’s map when they are not physically present). Example test performance data from an AMR is provided. The goal of this work is to provide a common method to characterize the performance of mobile systems in industrial environments that is easily comparable and communicated for both commercial and developmental purposes. This work is driven by existing