A Positioning Engine (PE) is the core software component responsible for computing position, velocity, time (PVT) solutions and associated quality metrics from raw GNSS observations, correction data, and sensor fusion inputs. This algorithmic heart of a positioning system implements the mathematical models and estimation techniques that transform measurements into actionable positioning information suitable for navigation, surveying, autonomous systems, and other applications.
The positioning engine performs multiple interconnected functions. It processes raw observations from the measurement engine, applying corrections for known error sources including satellite clock and orbit errors, atmospheric delays, and antenna phase center variations. It implements the position estimation algorithms, typically least squares or Kalman filtering approaches, that combine observations from multiple satellites to determine the receiver’s state. For RTK applications, it performs carrier phase ambiguity resolution to achieve centimeter-level accuracy. For integrated systems, it fuses GNSS with inertial, odometry, and potentially other sensor inputs to improve robustness and continuity.
Modern positioning engines for automotive and robotics applications must meet demanding requirements beyond basic accuracy. Integrity monitoring capabilities compute protection levels bounding worst-case position errors with specified probability, essential for safety-critical applications. Functional safety compliance (ISO 26262) ensures the software is developed and verified to appropriate safety standards. High update rates (10-100 Hz) support dynamic applications. Multi-constellation, multi-frequency processing maximizes accuracy and availability. Dead reckoning through GNSS outages maintains positioning continuity in tunnels or urban canyons.
Positioning engines may be implemented as software running on the receiver’s processor, on a separate host computer, or in cloud infrastructure. Software positioning engines offer flexibility, the same algorithm can run on different hardware platforms, and updates can be deployed without hardware changes. For safety-critical applications, the positioning engine architecture must address systematic factors including software quality, computational determinism, failure detection, and graceful degradation when inputs become unavailable or unreliable.
