Tight coupling of GPS, laser scanner, and inertial measurements for navigation in urban environments

Abstract

Many applications can be envisioned for accurate, robust, and reliable navigation solution in challenging urban environments. Examples of existing and prospective applications include, but are not limited to, navigation, guidance, and control of autonomous vehicles (including both ground and aerial vehicles) for urban surveillance and reconnaissance; collection of geographical information system (GIS) data in cities; monitoring of urban infrastructure for situational awareness; and, precise automotive applications such as automated lane keeping. If used by themselves, none of the existing navigation technologies have the potential to fully satisfy the requirements for reliable and accurate navigation in urban environments. Hence, this paper develops a multi-sensor integrated solution that combines the complementary features of the Global Positioning System (GPS), laser scanner feature-based navigation, and inertial navigation for urban scenarios. GPS and laser scanner-based navigation ideally complement each other for urban navigation. The laser scanner-based navigation relies on the availability of structures (lines and surfaces) within the scan range (80 m, typically). Features (such as lines) are first extracted from laser scan images and then used for position and attitude determination. In urban areas, if there exists a building wall that blocks GPS signals, this wall creates a feature in the laser scan image. On the other hand, for open streets with limited features, the GPS signal is generally unobstructed. Thus, GPS and laser data can be combined into integrated solution architecture. The system architecture developed also exploits INS navigation states for improved solution robustness: e.g., for robust feature association between scan images and for coasting through instances where sufficient number of combined GPS/laser measurements is unavailable. A tightly coupled GPS/laser scanner/INS mechanization is developed and applied for centimeter- accurate trajectory reconstruction. The paper uses live urban data to demonstrate that combined GPS and laser scanner data generally support the observability of navigation states at any part of the urban trajectory; and, for those limited cases where insufficient GPS/laser measurements are available, the INS coasting option can be efficiently utilized. Test results presented also show that the developed tightly coupled GPS/laser scanner/INS solution provides accurate trajectory reconstruction capabilities (one sigma error residuals are at a cm-level) in challenging urban environments.