In the race to improve autonomous systems, researchers in China have unveiled a new algorithm that accelerates navigation while maintaining precision. Called Optimized iSAM-Factor Graph Optimization (OiSAM-FGO), the method halves processing times compared with leading benchmarks yet delivers accuracy equal to state-of-the-art solutions.
Traditional navigation tools face persistent trade-of by Riko Seibo
Tokyo, Japan (SPX) Aug 22, 2025
In the race to improve autonomous systems, researchers in China have unveiled a new algorithm that accelerates navigation while maintaining precision. Called Optimized iSAM-Factor Graph Optimization (OiSAM-FGO), the method halves processing times compared with leading benchmarks yet delivers accuracy equal to state-of-the-art solutions.
Traditional navigation tools face persistent trade-offs. Global Navigation Satellite Systems (GNSS) struggle in cities, while Inertial Navigation Systems (INS) drift over time. Fusion algorithms like the Extended Kalman Filter help but fail to capture nonlinear dynamics. Factor Graph Optimization (FGO) emerged as a global solution but proved too computationally demanding for embedded devices.
The OiSAM-FGO framework, developed by the Institute of Microelectronics at the Chinese Academy of Sciences and the University of Chinese Academy of Sciences, addresses this challenge. It introduces an optimized incremental smoothing process that limits calculations to essential non-zero elements, reducing complexity from quadratic to linear scale. Adaptive Joint Sliding Window Re-linearization (A-JSWR) further streamlines performance by updating only when necessary.
Testing on the Awesome GINS dataset and field trials in Wuhan confirmed the algorithm's impact. Compared with OB-GINS, the current benchmark FGO method, and Extended Kalman Filters, OiSAM-FGO cut optimization time by more than 50% while sustaining superior accuracy in position, velocity, and attitude. Even under demanding conditions, robustness was maintained with minimal yaw fluctuations.
Lead author Zhichao Yang emphasized the broader implications: "With OiSAM-FGO, we've shown it is possible to retain the benefits of global optimization while stripping away much of the computational burden. This means resource-limited platforms-from embedded automotive systems to portable robotics-can now access levels of navigation accuracy once thought too expensive in terms of processing power."
The approach could extend to multi-sensor systems combining GNSS with LiDAR or visual data, advancing urban mobility and intelligent transport networks. By reducing power use and hardware costs, the new algorithm offers a practical path toward real-time, reliable navigation for autonomous vehicles, drones, and robotics.
Research Report:OiSAM-FGO: an efficient factor graph optimization algorithm for GNSS/INS integrated navigation system
Related Links
Aerospace Information Research Institute
GPS Applications, Technology and Suppliers