Control Barrier Corridors: From Safety Functions to Safe Sets
Authors: Ömür Arslan, Nikolay Atanasov
Organizations: Department of Mechanical Engineering, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands. · Department of Electrical and Computer Engineering, University of California San Diego, La Jolla, CA 92093, USA.
Safe autonomy is a critical requirement and a key enabler for robots to operate safely in unstructured complex environments. Control barrier functions and safe motion corridors are two widely used but technically distinct safety methods, functional and geometric, respectively, for safe motion planning and control. Control barrier functions are applied to the safety filtering of control inputs to limit the decay rate of system safety, whereas safe motion corridors are geometrically constructed to define a local safe zone around the system state for use in motion optimization and reference-governor design. This paper introduces a new notion of control barrier corridors, which unifies these two approaches by converting control barrier functions into local safe goal regions for reference goal selection in feedback control systems. We show, with examples on fully actuated systems, kinematic unicycles, and linear output regulation systems, that individual state safety can be extended locally over control barrier corridors for convex barrier functions, provided the control convergence rate matches the barrier decay rate, highlighting a trade-off between safety and reactiveness. Such safe control barrier corridors enable safely reachable persistent goal selection over continuously changing barrier corridors during system motion, which we demonstrate for verifiably safe and persistent path following in autonomous exploration of unknown environments.
Figures & tables
Fig. 1: Control barrier corridors turn CBF safety specification on control inputs into a geometric representation of safe goals for state feedback control. For instance, a control barrier corridor BC(x) around a safe state x of a fully actuated system x˙=u under proportional control u=−κ(x−x∗) is the set of goal states x∗ that ensure safe control based on the feasibility condition h˙i(x)≥−α(hi(x)) for a set of CBFs h1(x),…,hm(x) . (left) The control barrier corridor of a fully actuated circular robot, centered at x with radius r>0 , relative to a point obstacle at q , constructed based on the power-distance barrier function h(x)=∥x−q∥p−rp . The control barrier corridor enlarges with an increasing ratio of the barrier decay rate to the proportional control gain, κα , and shrinks with the increasing order p of the power distance. (right) The control barrier corridor (yellow) of a fully actuated circular robot (cyan+red arrow), constructed relative to multiple sensed obstacle points (red), captures a local collision-free space around the robot for convex barrier functions (e.g., the power distance with p≥1 ) under identical linear barrier decay rate and proportional control gain, α=κ . This allows for selecting a safely reachable goal (cyan) for following a reference path (blue).
Fig. 2: The influence of the proportional control gain κ , barrier decay rate α , and the order p of the power distance on control barrier corridors (yellow) of a fully actuated robot (cyan) relative to sensed obstacle points (red). A control barrier corridor corresponds to a local safe zone around the robot (cyan), free of obstacles (red), for a pair of matching barrier rate and control gain ( α=κ ), and convex barrier functions ( p≥1 ). The conservativeness of the barrier corridor increases with higher control gain ( α<κ ) and higher barrier convexity ( p>1 ), and it becomes less accurate with increasing barrier rate ( α>κ ) and increasing barrier concavity ( p<1 ).
Fig. 3: Control barrier corridor (yellow) of the soft minimum of multiple barrier functions based on the collision distance of a circular robot body (cyan) to sensed obstacle points (red), where the proportional gain and linear barrier decay rates are the same, i.e., α=κ . Due to the lack of convexity and approximation error, the soft-min barrier composition is unable to identify a collision-free convex neighborhood around the robot as a control barrier corridor. (left) λ=2 : small softness results in large approximation error and inaccurate safety assessment of the robot position. (middle) λ=10 : sufficiently large softness ensures robot safety but still inaccurately approximates the minimum distance to collision. (right) λ=100 : very large softness accurately approximates the minimum distance to collision, but its gradient becomes less smooth.
Fig. 4: Sensor-based control barrier corridors (yellow), constructed using (left) sensed unfree (red), i.e., occupied (black) and unknown(gray) locations in an occupancy grid map, (middle) the boundary (red) of the unfree space in the occupancy map, and (right) the sensed visible obstacle points from a laser scanner within a certain range around the robot, almost always result in the same barrier corridor, except for occlusions. Hence, visible obstacle points constitute the essential element for barrier safety and safe control.
Fig. 5: Local goal selection (cyan) for following a path (blue) using control barrier corridors (yellow). Matching the barrier decay rate and the control convergence rate ( α=κ , top) enables safe goal selection and reliable path-following control, as opposed to using a higher barrier decay rate than the control convergence rate ( α=3κ , bottom).
Fig. 6: Safe and persistent path following for frontier-based autonomous exploration of an unknown static environment, using a reference exploration path (blue line) toward a reachable frontier point (blue dot), with safely reachable path goal selection (cyan) over control barrier corridors (yellow) of a unicycle mobile robot (cyan + red arrow) constructed from sensed obstacle points (red) via an onboard laser scanner sensor.
As the number of autonomous robots continues to grow, safety becomes increasingly important. Control barrier functions (CBFs) provide a theoretically grounded framework for ensuring safety, but existing design methods often face limitations in effectiveness, scalability, or interpretability, and may result in overly conservative safe sets. In this paper, we propose VertexCBF, a framework for learning neural CBFs in a scalable, systematic, and explainable way. We approximate the stationary Hamilton--Jacobi value function using a neural network trained via a combination of physics-informed and sparsely supervised learning. By exploiting control-affine dynamics and a convex polytope control set, under which the Hamiltonian is maximized at the control vertices, we efficiently generate supervision points via GPU-parallel vertex-restricted tree search, while a residual architecture guarantees that the learned CBF is never larger than the specified constraint function. We evaluate the method on 15 systems and compare it against relevant baselines, showing that it reliably recovers large safe sets where the baselines are conservative or fail completely. In addition, we perform a hardware experiment in which a mobile robot safely avoids pedestrians using a neural CBF trained with our method.
Bojan Derajić, Sebastian Bernhard, Wolfgang Hönig
Technical University of Berlin, Germany · AUMOVIO, Germany · Technical University of Applied Sciences Augsburg, Germany +1
Autonomous mobile robots operating in tight environments require motion planning frameworks that account for the physical footprint of the robot. Simplifying the geometry to a point or a circle is conservative and discards information needed to successfully and safely traverse narrow passages. This work proposes a safe local motion planning and control method that guarantees that a polytopic robot footprint stays inside a continuously updated convex free-space region. The containment condition is formulated as a set of discrete-time control barrier function constraints within a model predictive controller. The number of safety constraints depends on the complexity of the local free-space geometry and the robot shape, instead of the number of obstacles. The proposed free-space formulation does not need any obstacle detection or segmentation. A comparative analysis against a polytope-based obstacle avoidance formulation confirms favorable scaling up to a reduction of 91× in computation time as the number of obstacles increases. The approach is validated in simulation with an autonomous surface vehicle and on hardware with a non-holonomic mobile robot, using both occupancy grids and LiDAR sensing. The experiments demonstrate safe real-time motion planning and control at 10~Hz on an onboard embedded computer, including reactive avoidance of dynamic obstacles.
Alejandro Gonzalez-Garcia, Dries Dirckx, Jan Swevers +1
This tutorial provides a critical review of the practical application of Control Barrier Functions (CBFs) in robotic safety. While the theoretical foundations of CBFs are well-established, I identify a recurring gap between the mathematical assumption of a safe controller's existence and its constructive realization in systems with input constraints. I highlight the distinction between candidate and valid CBFs by analyzing the interplay of system dynamics, actuation limits, and class-K functions. I further show that some purported demonstrations of safe robot policies or controllers are limited to passively safe systems, such as single integrators or kinematic manipulators, where safety is already inherited from the underlying physics and even naive geometric hard constraints suffice to prevent collisions. By revisiting simple low-dimensional examples, I show when CBF formulations provide valid safety guarantees and when they fail due to common misuses. I then provide practical guidelines for constructing realizable safety arguments for systems without such passive safety. A crowd-navigation simulation study further illustrates that CBF-derived reward shaping in reinforcement learning can improve empirical behavior without establishing formal safety. The goal of this tutorial is to bridge the gap between theoretical guarantees and actual implementation, supported by an open-source interactive web demonstration that visualizes these concepts intuitively.
Taekyung Kim
Department of Robotics, University of Michigan, Ann Arbor, MI, 48109, USA