TrajEdit: Real-Time Flow-Matching Trajectory Adaptation via FlowEdit and CBF Safety Filter

Wondmgezahu Teshome; Arya Honarpisheh; Octavia Camps; Mario Sznaier

cs.ROOct 16, 2026

TrajEdit: Real-Time Flow-Matching Trajectory Adaptation via FlowEdit and CBF Safety Filter

Authors: Wondmgezahu Teshome, Arya Honarpisheh, Octavia Camps, Mario Sznaier

Organizations: Department of Electrical and Computer Engineering, Northeastern University, Boston, MA, USA.

Abstract

Safe, real-time motion planning is especially difficult when a dynamic obstacle actively pursues the ego agent, requiring rapid adaptation as the environment changes. We present TrajEdit, a hierarchical trajectory planner for pursuit-evasion that combines flow matching, trajectory editing, and safety filtering. Rather than regenerating trajectories from scratch at every step, TrajEdit follows the current plan under nominal conditions and adapts its remaining portion through FlowEdit as the pursuer approaches, and falls back to a control barrier function safety filter if the adversarial agent is within a safety threshold. Our use of FlowEdit for trajectory editing reduces computation time while maintaining a comparable success rate. We evaluate TrajEdit in 2D pursuit-evasion with static obstacles and an adversarial pursuer, demonstrating high success rates with low capture and collision rates, while achieving computational time savings over full replanning baselines.

Figures & tables

Fig. 1: Overview of TrajEdit. (1) The scene and plan endpoints define the condition $\mathbf{c}$ . (2) The flow-matching backbone integrates $v_{\theta}$ from noise to a plan $\bm{\tau}$ . (3)–(5) At each step, the CBF filter acts if no plan is accepted or if the evader or its next waypoint is within $r_{\mathrm{esc}}$ of the pursuer, and the plan is then regenerated. Otherwise, the evader follows the plan, which is kept while the pursuer stays at least $r_{e}$ away and enough of the plan remains, and edited if not. In (4), dark blue marks the remaining plan, light blue its executed part, and grey a discarded plan; dashed arrows indicate feedback, $\bm{\tau}[k{+}1]$ is the next waypoint of $\bm{\tau}$ .

Fig. 2: Editing vs. replanning after two obstacles move. (a) Source plan on the original layout. (b) The edited plan stays close to the source, while replanning takes a different route. Circle: start; star: goal.

	Lead pursuit				Pure pursuit				Time (ms) $\downarrow$
Method	Goal(%) $\uparrow$	Cap.(%) $\downarrow$	Coll.(%) $\downarrow$	Abort(%) $\downarrow$	Goal(%) $\uparrow$	Cap.(%) $\downarrow$	Coll.(%) $\downarrow$	Abort(%) $\downarrow$	Lead / Pure
VO [ 7 ]	$49.40\pm 3.88$	$50.60\pm 3.88$	$0.00\pm 0.00$	–	$\mathbf{99.00}\pm 0.00$	$\underline{1.00}\pm 0.00$	$0.00\pm 0.00$	–	0.68 / 0.68
APF [ 5 ]	$5.00\pm 0.63$	$92.00\pm 1.79$	$0.00\pm 0.00$	–	$74.40\pm 3.20$	$13.80\pm 2.32$	$0.00\pm 0.00$	–	0.13 / 0.13
TD3 [ 30 ]	$54.20\pm 1.94$	$18.40\pm 4.22$	$23.80\pm 2.14$	–	$85.20\pm 4.45$	$5.40\pm 2.65$	$9.00\pm 2.19$	–	0.25 / 0.26
SAC [ 29 ]	$\underline{57.00}\pm 3.03$	$\underline{16.80}\pm 2.32$	$25.20\pm 2.48$	–	$84.80\pm 2.14$	$5.40\pm 1.85$	$9.60\pm 2.33$	–	0.63 / 0.63
RAMP [ 16 ]	$33.40\pm 3.50$	$50.20\pm 5.04$	$0.00\pm 0.00$	$15.00\pm 2.76$	$70.40\pm 3.61$	$25.20\pm 3.31$	$0.00\pm 0.00$	$4.40\pm 1.36$	528.87 / 495.85

TABLE I: Comparison with baselines under lead and pure pursuit on the convex layout. Best in bold, second best underlined (Goal and Cap.), – : not applicable

Lead pursuit
Variant	Goal(%) $\uparrow$	Abort(%) $\downarrow$	Time (ms) $\downarrow$
TrajEdit	$83.00\pm 2.76$	$16.00\pm 3.22$	$53.65$
w/o FlowEdit	$81.80\pm 3.54$	$17.20\pm 3.31$	$65.63$
w/o planner	$67.60\pm 5.54$	$27.20\pm 5.23$	$1.92$
Pure pursuit
TrajEdit	$96.80\pm 1.47$	$2.80\pm 1.72$	$47.11$

TABLE II: Ablation on the convex layout

Lead pursuit
Variant	Goal(%) $\uparrow$	Abort(%) $\downarrow$	Time (ms) $\downarrow$
TrajEdit	$51.60\pm 6.18$	$43.60\pm 3.98$	$54.54$
w/o FlowEdit	$55.80\pm 4.26$	$38.00\pm 2.76$	$66.75$
w/o planner	$49.80\pm 5.15$	$44.80\pm 4.53$	$1.95$
Pure pursuit
TrajEdit	$75.80\pm 2.79$	$24.00\pm 2.53$	$48.52$

TABLE III: Ablation on the concave layout

Fig. 3: Failure rate under local perturbations of one and two static obstacles, across perturbation radius.

	Goal (%)	Abort (%)	Cap. (%)
Per-step, $r_{\mathrm{cap}}$	$13.20\pm 1.60$	$4.60\pm 2.24$	$82.20\pm 3.25$
Gated, $r_{\mathrm{cap}}$	$18.00\pm 3.03$	$3.80\pm 2.04$	$78.20\pm 3.19$
Per-step, $r_{\mathrm{cap}}+v^{p}_{\max}\Delta_{t}$	$61.00\pm 3.63$	$18.20\pm 1.33$	$19.20\pm 2.48$
Gated, $r_{\mathrm{cap}}+v^{p}_{\max}\Delta_{t}$	$68.60\pm 5.43$	$12.80\pm 2.48$	$17.60\pm 5.85$

TABLE IV: Sensitivity of SFM to the pursuer barrier radius (pure pursuit)

Safe Motion Planning

Safe motion planning develops methods for robots and vehicles to reach goals while avoiding collisions and respecting dynamic and uncertainty constraints.

See all 215 papers

Finite-Time Capture by a Dubins Pursuer Against Unknown Evader Strategies
Oct 18, 2026Dinesh Patra, Bhargav Jha, Ashish R. HotaFeedback Control
Indian Institute of Technology Kharagpur, Kharagpur, West Bengal 721302, India
Adaptive Risk-Certified Event-Triggered Replanning for Dynamic Navigation
Oct 7, 2026Richie R. Suganda, Bin HuRisk-Aware Path Planning Safe Motion Planning
Department of Electrical Engineering and Computer University of Houston Houston, TX 77004, USA · Department of Engineering Technology University of Houston Houston, TX 77004, USA
STITCHER: Constrained Trajectory Planning in Complex Environments with Real-Time Motion Primitive Search
Oct 16, 2025Helene J. Levy, Brett T. LopezConstrained Motion Planning Trajectory Generation
VECTR Laboratory, University of California, Los Angeles, Los Angeles, CA, USA

Oct 18, 2026J/K move · Enter open · S save

Finite-Time Capture by a Dubins Pursuer Against Unknown Evader Strategies

Dinesh Patra, Bhargav Jha, Ashish R. Hota

Indian Institute of Technology Kharagpur, Kharagpur, West Bengal 721302, India

Feedback Control

Open paper AskPDF arXiv

We consider a planar pursuit-evasion problem in which a Dubins pursuer seeks to capture a slower first-order point-mass evader that can change its heading instantaneously. First, we propose a turn-rate feedback law with a single tunable gain and derive an explicit admissible gain interval that guarantees finite-time capture against any admissible evader strategy while satisfying the turn-rate constraint. Next, we obtain an explicit capture-time bound and characterize the turning authority required for the gain interval to be non-empty. Finally, we show that the dependence on the evader's speed bound can be removed by deriving a speed ceiling, computable solely from the pursuer's parameters, below which the guarantees hold at the maximum allowable gain. Numerical simulations illustrate the effectiveness of the proposed approach.

Abstract

Safe Motion Planning

Explore similar work

Finite-Time Capture by a Dubins Pursuer Against Unknown Evader Strategies

Adaptive Risk-Certified Event-Triggered Replanning for Dynamic Navigation

STITCHER: Constrained Trajectory Planning in Complex Environments with Real-Time Motion Primitive Search