Guidance vector fields (GVFs) are a popular framework for tasks such as robotic path-following and multi-robot motion coordination, but existing approaches may be non-differentiable or have large derivatives, limiting their use in controllers that require derivatives of the guidance signal, which is common when the GVF is used as an input in a cascaded controller. This paper proposes an infinitely differentiable constructive GVF. The resulting vector field depends on a smoothing parameter and is a generalization of [1], recovering the original GVF in the limit as this parameter approaches zero. It also admits a simple and efficient implementation that preserves smoothness under discretization and can reduce computation time by using substantially fewer curve samples while maintaining small path-following errors. We establish the main theoretical properties and demonstrate its practical advantages through numerical and real-world experiments.
Figures & tables
Fig. 1 : Illustration of the discontinuity problem in [ 1 ] . The target curve is the ellipse, and the red line segment represents its medial axis, where the distance function is nondifferentiable. The plots on the right show the two components of the GVF in [ 1 ] (orange) and of the proposed GVF (blue), evaluated along the magenta line. Note the discontinuity of the orange vector field at p2=0 , which lies on the medial axis.
Fig. 2 : Drone used in the experiment and simulation
Name
Value
Name
Value
Name
Value
G
2.0s−1
H
1.0m/s
Kv
2.0s−1
KΔ
2.0s−1
Kω
2.0s−1
M
2.09kg
Δt
0.001s
Tmin
0N
Tmax
30N
TABLE I : Parameter values (simulation and experiment).
Fig. 3 : Computational-budget experiment using five target curves and ten initial positions per curve. Top: average steady-state tracking error among successful trajectories. Middle: failure rate. Bottom: best-performing tested smoothing parameter r>0 . Each implementation uses the largest number of segments estimated to fit within the evaluation budget.
Fig. 4 : Comparison of the control inputs from the GVF and their numerical derivatives considering different implementations.
Fig. 5 : Target path (black) and tracking trajectory (green) in the flight experiment
Fig. 6 : Curve tracking error 2DC(p(t)) in the experiment, in relation to the original curve C .
Fig. 7 : Physical measured states in the experiment.
Fig. 8 : Control inputs in the experiment.
Fig. 9 : Auxiliary variables and virtual state vd
Guiding Vector Fields (GVFs) are a powerful tool for robotic path following. However, classical methods assume smooth, ordered curves and fail when paths are unordered, multi-branch, or generated by probabilistic models. We propose a unified framework, termed the Score-Induced Guiding Vector Field (SGVF), which leverages score-based generative modeling to construct vector fields directly from data distributions. SGVF learns tangent fields from point clouds with unit-norm, orthogonality, and directional-consistency losses, ensuring geometric fidelity and control feasibility. This approach removes the reliance on ad-hoc path segmentation and enables guidance along complex topologies such as branching and pseudo-manifolds. The study establishes a correspondence between score vanishing in diffusion models and GVF singularities and highlights representational capacity near sharp path curvatures. Experiments on robotic navigation in planar environments demonstrate that SGVF achieves reliable path following in scenarios where classical GVFs fail, underscoring its potential as a bridge between generative modeling and geometric control. Code and experiment video are available at https://github.com/czr-gif/Guiding-Vector-Field-Generation-via-Score-based-Diffusion-Model.
Zirui Chen, Shiliang Guo, Shiyu Zhao
College of Computer Science and Technology, Zhejiang University, Hangzhou, China. · WINDY Lab, Department of Artificial Intelligence, Westlake University, Hangzhou, China.
The existing singularity-free guiding vector field (SF-GVF) with an additional virtual coordinate can eliminate singular points (i.e., points where the vector field vanishes) inherent in conventional GVFs and guarantee global convergence of robot trajectories to closed and self-intersecting desired paths. However, the desired speed given by the GVF along the desired path in the original lower-dimensional space cannot be arbitrarily specified but depends on path parameterizations. One possible workaround is to partially normalize the physical projection of the SF-GVF and assign a user-designed speed. However, we show that this workaround may introduce new singularities since the normalization denominator can become zero. To address this issue, we propose a new SF-GVF with prescribed physical speed (PPS). The integral curves of the new SF-GVF converge exponentially to the desired path from any initial condition in the higher-dimensional space (including virtual dimension); more importantly, the robot's physical speed converges to the PPS, while the path-error dynamics remain invariant under regular reparameterizations of the desired path. We further develop a saturated acceleration control law for second-order kinematic models. Finally, comparative simulations and 3D path-following experiments with a quadrotor under different PPS profiles validate the theoretical results and demonstrate the effectiveness of the proposed approach.
Zhouru Xiao, Sha Luo, Yang Lu +5
School of Artificial Intelligence and Robotics, Hunan University, Changsha, China · College of Information Science and Engineering, Hunan Normal University, Changsha, China · College of Systems Engineering and the State Key Laboratory of Digital Intelligent Modeling and Simulation, National University of Defense Technology, Changsha, China +2
This paper develops a singularity-free guiding vector field (SF-GVF) for path following on the special orthogonal group SO(3). First, we lift the Euclidean SF-GVF construction to SO(3), integrating the augmented-state approach with the intrinsic Lie-group geometry and obtaining a closed-form geometric guidance law whose integral curves converge to a designer-specified attitude path. The field is defined on a dense open subset of SO(3), excluding only the measure-zero antipodal set - a manifestation of the topological obstruction to continuous global stabilization on SO(3). The construction requires no per-step optimization and produces a control input intrinsically in so(3) as body angular rates. Second, we formalize the progression behavior along the path as a designer-supplied function ν(ξ), promoting the parametric speed from an implicitly resolved degree of freedom to a first-class design specification. In contrast to the Euclidean condition v = 0, which excludes vehicles with minimum-speed constraints, the corresponding condition ω= 0 on SO(3) is physically admissible for most platforms with active attitude control, making the progression behavior a design freedom structurally available on SO(3) but absent in the Euclidean setting. The framework's structural results are established under a bi-invariant Riemannian metric and hold uniformly across choices of path, progression, and Lyapunov gain. The framework is illustrated in simulation on self-intersecting paths under both constant and point-convergence progression behaviors.
Jesus Bautista, Hector Garcia de Marina
Department of Computer Engineering, Automation and Robotics, and Institute of Mathematics (IMAG), University of Granada, Granada, Spain