Inside Unmanned Systems provides actionable business intelligence to decision-makers and influencers operating within the global UAS community. Features include analysis of key technologies, policy/regulatory developments and new product design.
Issue link: https://insideunmanned.epubxp.com/i/668560
64 unmanned systems inside April/May 2016 GROUND AUTOMOTIVE was developed and implemented. This setup is designed to include a variety of sensors and sup- port development and demonstration efforts for various use cases and accuracy requirements from centimeter-level urban positioning for surveying and mapping to meter-level localiza- tion for automotive safety. Figure 9 shows the ground vehicle test setup that includes GPS receivers (survey-grade No- vAtel and consumer-grade NVS receivers with pinwheel and patch antennas); a higher-grade MEMS IMU (~10 deg/hr, STIM-300 manu- factured by Sensonor); a PX4 autopilot with consumer-grade IMU (~100 deg/hr unit manu- factured by ST Micro), magnetometer and baro- altimeter; Prosilica video cameras, SICK LMS- 200 scanning lidar, Microsemi chip-scale atomic clock (CSAC); data synchronization and data collection units. The physical layout of the data collection system has been designed to support the real-time demonstration objective. Figure 10 shows the annotated sensor board. The data collection and processing system has been pal- letized and mounted in a transportable 19"-6U equipment case as shown in Figure 11. The test setup was utilized to evaluate various low-cost solutions for automotive safety applications. For experimental demonstrations, experi- mental data were collected in urban canyons of downtown San Francisco, CA in January 2016. Figure 12 illustrates typical test environments. First, we evaluated GNSS-only performance. As expected, GNSS positioning capabilities were found to be extremely limited. Figure 13 shows GNSS (GPS+GLONASS) position solution dis- played in Google Earth. Sparse and unreliable position fixes are obtained, which is clearly un- satisfactory for automotive safety applications. Next, PnP navigation software was auto- matically reconfigured to evaluate GNSS/INS integration with consumer-grade inertial. Car- rier phase measurements were processed using temporal phase differences (to eliminate integer ambiguities) and relative position observables of RIFE. Figures 14 and 15 show example test Figure 13: Typical performance of GNSS position solution in downtown environments Figure 14: Position solution of carrier phase GNSS integrated with consumer-grade inertial for example test scenario 1: consitnious trajectory reconstruction is obtained; however, in dense urban canyons signif cant deviations can be present (as indicated in the zoomed image in the bottom) Figure 15: Position solution of carrier phase GNSS integrated with consumer-grade inertial for example test scenario 2