About

Professor Steven M. LaValle is primarily interested in the design of planning algorithms, focusing on problems involving continuous spaces, complex geometric constraints, differential constraints, and sensing uncertainties. His research addresses fundamental issues in robotics, computer graphics, architectural design, and computational biology. His work builds on literature from robotics, algorithms, computational geometry, artificial intelligence, and control theory, combining theoretical analysis with practical implementation. He developed the Motion Strategy Library (MSL) with his students to promote the use of planning algorithms in research, education, and industry.

Research topics

• Computer Science
• Human–computer interaction
• Psychology
• Artificial Intelligence
• Multimedia
• Engineering
• Computer graphics (images)
• Psychotherapist
• Social psychology
• Applied psychology
• Computer vision
• Simulation

Selected publications

• Relating Reinforcement Learning to Dynamic Programming-Based Planning

  arXiv (Cornell University) · 2026-03-08

  articleOpen accessSenior author

  This paper bridges some of the gap between optimal planning and reinforcement learning (RL), both of which share roots in dynamic programming applied to sequential decision making or optimal control. Whereas planning typically favors deterministic models, goal termination, and cost minimization, RL tends to favor stochastic models, infinite-horizon discounting, and reward maximization in addition to learning-related parameters such as the learning rate and greediness factor. A derandomized version of RL is developed, analyzed, and implemented to yield performance comparisons with value iteration and Dijkstra's algorithm using simple planning models. Next, mathematical analysis shows: 1) conditions under which cost minimization and reward maximization are equivalent, 2) conditions for equivalence of single-shot goal termination and infinite-horizon episodic learning, and 3) conditions under which discounting causes goal achievement to fail. The paper then advocates for defining and optimizing truecost, rather than inserting arbitrary parameters to guide operations. Performance studies are then extended to the stochastic case, using planning-oriented criteria and comparing value iteration to RL with learning rates and greediness factors.

  PublisherOA PDF

• Smooth Feedback Motion Planning With Reduced Curvature

  IEEE Robotics and Automation Letters · 2026-04-09

  articleOpen accessSenior author

  Feedback motion planning over cell decompositions provides a robust method for generating collision-free robot motion with formal guarantees. However, existing algorithms often produce paths with unnecessary bending, leading to slower motion and higher control effort. This paper presents a computationally efficient method to mitigate this issue for a given simplicial decomposition. A heuristic is introduced that systematically aligns and assigns local vector fields to produce more direct trajectories, complemented by a novel geometric algorithm that constructs a maximal star-shaped chain of simplexes around the goal. This creates a large “funnel” in which an optimal, direct-to-goal control law can be safely applied. Simulations demonstrate that our method generates measurably more direct paths, reducing total bending by an average of 91.40% and LQR control effort by an average of 45.47%. Furthermore, comparative analysis against sampling-based and optimization-based planners confirms the time efficacy and robustness of our approach. While the proposed algorithms work over any finite-dimensional simplicial complex embedded in the collision-free subset of the configuration space, the practical application focuses on low-dimensional (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$d\le 3$</tex-math></inline-formula>) configuration spaces, where simplicial decomposition is computationally tractable.

  PublisherDOI

• Planning Smooth and Safe Control Laws for a Unicycle Robot Among Obstacles

  arXiv (Cornell University) · 2026-04-19

  preprintOpen accessSenior author

  This paper presents a framework for safe navigation of a unicycle point robot to a goal position in an environment populated with obstacles from almost any admissible state, considering input limits. We introduce a novel QP formulation to create a Cinfinity-smooth vector field with reduced total bending and total turning. Then we design an analytic, non-linear feedback controller that inherently satisfies the conditions of Nagumo's theorem, ensuring forward invariance of the safe set without requiring any online optimization. We have demonstrated that our controller, even under hard input limits, safely converges to the goal position. Simulations confirm the effectiveness of the proposed framework, resulting in a twice faster arrival time with over 50\% lower angular control effort compared to the baseline.

  PublisherDOI

• Localization with Single or Antipodal Distance Measurements

  Springer proceedings in advanced robotics · 2026-01-01

  book-chapter
  PublisherDOI

• Mobile Robot Localization Using a Novel Whisker-Like Sensor

  arXiv (Cornell University) · 2026-01-09

  preprintOpen access

  Whisker-like touch sensors offer unique advantages for short-range perception in environments where visual and long-range sensing are unreliable, such as confined, cluttered, or low-visibility settings. This paper presents a framework for estimating contact points and robot localization in a known planar environment using a single whisker sensor. We develop a family of virtual sensor models. Each model maps robot configurations to sensor observations and enables structured reasoning through the concept of preimages - the set of robot states consistent with a given observation. The notion of virtual sensor models serves as an abstraction to reason about state uncertainty without dependence on physical implementation. By combining sensor observations with a motion model, we estimate the contact point. Iterative estimation then enables reconstruction of obstacle boundaries. Furthermore, intersecting states inferred from current observations with forward-projected states from previous steps allow accurate robot localization without relying on vision or external systems. The framework supports both deterministic and possibilistic formulations and is validated through simulation and physical experiments using a low-cost, 3D printed, Hall-effect-based whisker sensor. Results demonstrate accurate contact estimation and localization with errors under 7 mm, demonstrating the potential of whisker-based sensing as a lightweight, adaptable complement to vision-based navigation.

  PublisherDOI

• Planning Smooth and Safe Control Laws for a Unicycle Robot Among Obstacles

  arXiv (Cornell University) · 2026-04-19

  articleOpen accessSenior author

  This paper presents a framework for safe navigation of a unicycle point robot to a goal position in an environment populated with obstacles from almost any admissible state, considering input limits. We introduce a novel QP formulation to create a Cinfinity-smooth vector field with reduced total bending and total turning. Then we design an analytic, non-linear feedback controller that inherently satisfies the conditions of Nagumo's theorem, ensuring forward invariance of the safe set without requiring any online optimization. We have demonstrated that our controller, even under hard input limits, safely converges to the goal position. Simulations confirm the effectiveness of the proposed framework, resulting in a twice faster arrival time with over 50\% lower angular control effort compared to the baseline.

  PublisherOA PDF

• Smooth Feedback Motion Planning with Reduced Curvature

  arXiv (Cornell University) · 2026-04-02

  preprintOpen accessSenior author

  Feedback motion planning over cell decompositions provides a robust method for generating collision-free robot motion with formal guarantees. However, existing algorithms often produce paths with unnecessary bending, leading to slower motion and higher control effort. This paper presents a computationally efficient method to mitigate this issue for a given simplicial decomposition. A heuristic is introduced that systematically aligns and assigns local vector fields to produce more direct trajectories, complemented by a novel geometric algorithm that constructs a maximal star-shaped chain of simplexes around the goal. This creates a large ``funnel'' in which an optimal, direct-to-goal control law can be safely applied. Simulations demonstrate that our method generates measurably more direct paths, reducing total bending by an average of 91.40\% and LQR control effort by an average of 45.47\%. Furthermore, comparative analysis against sampling-based and optimization-based planners confirms the time efficacy and robustness of our approach. While the proposed algorithms work over any finite-dimensional simplicial complex embedded in the collision-free subset of the configuration space, the practical application focuses on low-dimensional ($d\le3$) configuration spaces, where simplicial decomposition is computationally tractable.

  PublisherDOI

• Mobile Robot Localization Using a Novel Whisker-Like Sensor

  ArXiv.org · 2026-01-09

  articleOpen access

  Whisker-like touch sensors offer unique advantages for short-range perception in environments where visual and long-range sensing are unreliable, such as confined, cluttered, or low-visibility settings. This paper presents a framework for estimating contact points and robot localization in a known planar environment using a single whisker sensor. We develop a family of virtual sensor models. Each model maps robot configurations to sensor observations and enables structured reasoning through the concept of preimages - the set of robot states consistent with a given observation. The notion of virtual sensor models serves as an abstraction to reason about state uncertainty without dependence on physical implementation. By combining sensor observations with a motion model, we estimate the contact point. Iterative estimation then enables reconstruction of obstacle boundaries. Furthermore, intersecting states inferred from current observations with forward-projected states from previous steps allow accurate robot localization without relying on vision or external systems. The framework supports both deterministic and possibilistic formulations and is validated through simulation and physical experiments using a low-cost, 3D printed, Hall-effect-based whisker sensor. Results demonstrate accurate contact estimation and localization with errors under 7 mm, demonstrating the potential of whisker-based sensing as a lightweight, adaptable complement to vision-based navigation.

  PublisherOA PDF

• Smooth Feedback Motion Planning with Reduced Curvature

  ArXiv.org · 2026-04-02

  articleOpen accessSenior author

  Feedback motion planning over cell decompositions provides a robust method for generating collision-free robot motion with formal guarantees. However, existing algorithms often produce paths with unnecessary bending, leading to slower motion and higher control effort. This paper presents a computationally efficient method to mitigate this issue for a given simplicial decomposition. A heuristic is introduced that systematically aligns and assigns local vector fields to produce more direct trajectories, complemented by a novel geometric algorithm that constructs a maximal star-shaped chain of simplexes around the goal. This creates a large ``funnel'' in which an optimal, direct-to-goal control law can be safely applied. Simulations demonstrate that our method generates measurably more direct paths, reducing total bending by an average of 91.40\% and LQR control effort by an average of 45.47\%. Furthermore, comparative analysis against sampling-based and optimization-based planners confirms the time efficacy and robustness of our approach. While the proposed algorithms work over any finite-dimensional simplicial complex embedded in the collision-free subset of the configuration space, the practical application focuses on low-dimensional ($d\le3$) configuration spaces, where simplicial decomposition is computationally tractable.

  PublisherOA PDF

• Minimally Sufficient Structures for Information-Feedback Policies

  Springer proceedings in advanced robotics · 2026-01-01

  preprintOpen accessSenior author
  PublisherOA PDFDOI

Recent grants

• RI: Medium Collaborative Research: Minimalist Mapping and Monitoring

  NSF · $432k · 2009–2013

• Expanding the Frontiers of Motion Planning: Feedback, Differential Constraints, and Resolution Completeness

  NSF · $351k · 2006–2010

• CPS: Small: Sensor Lattices

  NSF · $490k · 2010–2016

• NRI: Large: Collaborative Research: Human-robot Coordinated Manipulation and Transportation of Large Objects

  NSF · $300k · 2013–2019

Frequent coauthors

• Katherine J. Mimnaugh

  35 shared

• Basak Sakcak

  University of Oulu

  29 shared

• Markku Suomalainen

  VTT Technical Research Centre of Finland

  29 shared

• Israel Becerra

  Mathematics Research Center

  27 shared

• Jingjin Yu

  25 shared

• Seth Hutchinson

  21 shared

• Rafael Murrieta-Cid

  Mathematics Research Center

  21 shared

• Timo Ojala

  University of Oulu

  19 shared

Education

• Ph.D., Computer Science

  University of California, Berkeley

  1995

• M.S., Computer Science

  University of California, Berkeley

  1992

• B.S., Electrical Engineering and Computer Science

  University of California, Berkeley

  1990