16 correct drift, but no longer define the motion estimate. It is the key distinction between a GNSS‑reliant system and one built for resilience. A Growing Ecosystem of Complimentary External Aiding Sensors As autonomous platforms diversify, so do the sensors that support them. Gallup noted that engineers now commonly use “visual odometry and simultaneous localisation and mapping (SLAM), light detection and ranging (LiDAR) odometry, wheel encoders on ground vehicles, barometric altitude sensors, magnetometers, Doppler velocity sensors, and high-performance external GNSS receivers used selectively.” Each sensor contributes something different: • LiDAR provides strong geometric structure for mapping and motion. • Cameras supply rich visual cues within a structured environment. • Wheel encoders stabilize short‑term motion on ground vehicles. • Doppler sensors improve velocity estimates over the ground. • Barometers help anchor altitude. The goal is not to choose one sensor, but to fuse many. “When you do that well,” Gallup said, “you get a meaningful reduction in drift and much better endurance through GNSS outages.” How Fusion Engines Make the System Work Most embedded platforms use a loosely coupled navigation architecture. External systems—whether LiDAR, SLAM, or a standalone GNSS receiver—compute their own solutions. Those outputs are then injected into the inertial Extended Kalman Filter (EKF). An EKF is a math engine that constantly blends sensor data—like GNSS, IMUs, LiDAR, or wheel encoders—to produce one stable, best‑guess position for the vehicle even when individual sensors get noisy or unreliable. Maybe a better description of an EKF is that it works like a group of hired farm hands comparing notes in the field where the IMU, GNSS, LiDAR, and wheel sensors each tell their version of where the machine is. The EKF sorts through all those stories, throws out the ones that don’t make sense, and gives the vehicle one clean, reliable answer so it stays on the right line. But the real engineering challenge lies in weighting and validating those measurements. “Not all measurements are equally trustworthy,” Gallup explained. While a LiDAR solution in a clean warehouse may be excellent; in a dusty tunnel, it may be unusable. The fusion engine must constantly evaluate reliability and adjust weighting accordingly. A strong inertial core slows drift between updates, giving the system more breathing room when external sensors become unreliable. “Better inertial performance translates directly into better resilience,” she said. Why Timing Matters More Than Most Engineers Expect One of the most overlooked challenges in performing navigation, is timing. Gallup called it “critical, and consistently underestimated.” In an example, she said that if a vehicle is moving at 15 meters per second (49.2 feet/second), even a 1‑millisecond (0.001-second) timing error can cause significant spatial misalignment—especially when aligning LiDAR point clouds with inertial data. High‑performance systems therefore rely on microsecond‑level timestamp accuracy, pulse‑per‑second (PPS) synchronization, and low‑latency data paths. Without tight timing, even excellent sensors can produce poor results. Managing GNSS Degradation, Denial, and Reacquisition A resilient system does not treat GNSS as simply on or off. Instead, it manages through phases: available, degraded, denied, and reacquisition. When GNSS degrades, uncertainty increases and the filter shifts weight toward inertial propagation and other aiding sources. In full denial—such as an RF‑jammed tunnel—the system runs entirely on dead reckoning. Drift grows, but with good bias stability and calibration, it remains manageable.
RkJQdWJsaXNoZXIy MTQyMDk=