HIT and Politecnico di Torino develop robust powered descent guidance under mass uncertainty

Pinpoint soft landing—whether for Mars landers or reusable launch vehicles—depends critically on powered descent, when the vehicle must rely on engine thrust to decelerate and precisely shape its trajectory. However, a persistent, mission-critical obstacle is mass uncertainty: fueling uncertainty, residual propellant errors, and variations in fuel consumption can change the thrust-to-weight ratio, amplifying trajectory deviations and increasing the risk of constraint violations (e.g., terrain clearance). In other words, even an “optimal” nominal plan can become unsafe once real mass differs from assumptions, which is why a guidance method that remains reliable under uncertainty is essential for safe, autonomous landing.

Against this backdrop, a new collaboration between Harbin Institute of Technology (HIT), HIT (Shenzhen), and Politecnico di Torino presents a robust powered descent guidance framework that explicitly accounts for mass and fuel-consumption uncertainties while preserving the computational efficiency of convex optimization—a practical requirement for onboard, real-time trajectory generation.

First, the team models how mass uncertainty affects powered descent by translating thrust-to-mass variations into an effective acceleration error. Under their assumptions, the resulting position deviations are bounded and can be treated in an axis-decoupled way, which keeps the problem tractable for real-time convex optimization. Building on this, they address the worst-case descent scenario—the key point of the work—by ensuring the glide-slope (terrain-avoidance) constraint remains feasible throughout powered descent. Using geometric insight and theoretical analysis, they derive a practical robust constraint-tightening strategy, so the planned trajectory stays safe and constraint-satisfying even when actual mass deviates from the nominal value.

To turn robust planning into reliable flight behavior, the framework further integrates a receding-horizon “plan–execute–replan” closed-loop guidance scheme that repeatedly updates the onboard solution, limiting error accumulation and improving terminal accuracy. Simulations show that both nominal and disturbed trajectories under mass uncertainty satisfy path constraints and avoid ground collision. Compared with a probabilistic robust method based on polynomial chaos expansion (PCE), the proposed worst-case approach can track constraint boundaries more tightly (reducing unnecessary conservatism), while PCE-based methods may need extra margins that can increase propellant usage. Meanwhile, by preserving the convex structure without high-order state expansion, the method maintains low computational burden and remains well-suited to real-time guidance.

Looking forward, the authors note that the framework can be extended beyond mass uncertainty to broader uncertainty sources, such as navigation errors in position and velocity. The next logical step is therefore to incorporate these additional uncertainties into the same robust-and-efficient guidance architecture and to further validate performance across wider disturbance sets and mission profiles. Ultimately, the goal is a more general, onboard-ready foundation for autonomous landing that can maintain safety guarantees and high accuracy in complex, uncertain environments—supporting future planetary landings and reusable launch operations.

Original Source

Duozhi GAO, Yanning GUO, Edoardo FADDA, Youmin GONG, Chuanjiang LI, Paolo BRANDIMARTE. Robust powered descent guidance considering mass and fuel consumption uncertainties: A convex optimization approach[J]. Chinese Journal of Aeronautics, 2026, https://doi.org/10.1016/j.cja.2025.103914.

About Chinese Journal of Aeronautics 

Chinese Journal of Aeronautics (CJA) is an open access, peer-reviewed international journal covering all aspects of aerospace engineering, monthly published by Elsevier. The Journal reports the scientific and technological achievements and frontiers in aeronautic engineering and astronautic engineering, in both theory and practice. CJA is indexed in SCI (IF = 5.7, Q1), EI, IAA, AJ, CSA, Scopus.