Logic-Geometric Programming: An Optimization-Based Approach to Combined Task and Motion Planning

By Marc Toussaint. 2015

372 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Reinforcement Learning

Citations in the corpus

Do As I Can, Not As I Say: Grounding Language in Robotic Affordances (2022, background): “Classically, this problem has been solved through symbolic planning [61, 63] or optimization [65, 66], but these require explicit primitives and constraints.”
Inner Monologue: Embodied Reasoning through Planning with Language Models (2022, background): “Traditionally, this problem has been solved through optimization [28, 29] or symbolic reasoning [24, 26], but more recently machine learning has been applied to aspects of the problem via learned representations, learned task-primitives, and more [30, 31, 32, 33, 34, 35, 36, 37, 38].”
Structured agents for physical construction (2019, background): “, 2012), and logic-based optimization for open-ended objectives in robotics (Toussaint, 2015).”
Differentiable Physics and Stable Modes for Tool-Use and Manipulation Planning (2018, methodology): “We call this a Logic-Geometric Program [ Tou-ssaint, 2015; Toussaint and Lopes, 2017; Toussaint et al. , 2018 ] , Here, ˆ x = ( x, ˙ x, ˙ x (cid:48) ) captures the conﬁguration velocity ˙ x before and after ( ˙ x (cid:48) ) an instantaneous impulse exchange at a state transition.”
Text2Motion: from natural language instructions to feasible plans (2023, background): “While different TAMP methods have relaxed the dependence on these components, many prominent works require all three [32, 36, 53, 12, 4, 20].”
Physical problem solving: Joint planning with symbolic, geometric, and dynamic constraints (2017, methodology): “Due to space limits, we cannot provide any further the details of KOMO and the logic-geometric programming framework (but see Toussaint, 2015, 2014).”
Deep Visual Reasoning: Learning to Predict Action Sequences for Task and Motion Planning from an Initial Scene Image (2020, methodology): “This work relies on Logic Geometric Programming (LGP) [42, 43] as the underlying TAMP framework.”
ReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic Manipulation (2024, background): “Logic-Geometric Programming [81–86] alternatively formulates a nonlinear constrained program over the entire state trajectory, taking into account both logical and geometric constraints.”
Towards Practical Multi-Object Manipulation using Relational Reinforcement Learning (2019, result): “Other lines of work [52], [53] achieved impressive results on stacking objects, but relied extensively”
Broadly-Exploring, Local-Policy Trees for Long-Horizon Task Planning (2020, methodology): “As the temporal distance is not easily quantifiable, we learn to predict it with a temporal distance classifier, ⌧tdc(xi, xj), with K exponential categories corresponding to temporal distance intervals, dk 2 {[0, 1), [1, 2), [2, 4), [4, 8), [8, 16), .”
PDDLStream: Integrating Symbolic Planners and Blackbox Samplers via Optimistic Adaptive Planning (2018, background): “…approaches to robotic task and motion planning have developed strategies for handling continuous spaces that go beyond prediscretization (Kaelbling and LozanoPérez 2011; Srivastava et al. 2014; Garrett, Lozano-Pérez, and Kaelbling 2015; Toussaint 2015; Garrett, Lozano-Pérez, and Kaelbling 2017).”
Integrated Task and Motion Planning (2020, background): “Some approaches even solve relaxations of the induced h-csp s that omit certain constraints such as motion constraints with many decision variables, which are often satisﬁable and thus are uninformative (65, 66, 67, 61, 68, 69).”
Visual Robot Task Planning (2018, background): “In robotics, effective TAMP approaches have been developed for solving complex problems involving spatial reasoning [11].”
Hierarchical Planning for Long-Horizon Manipulation with Geometric and Symbolic Scene Graphs (2020, methodology): “They perform discrete task plans with logic search and continuous motion plans through sampling-based techniques [10, 16] or trajectory optimization [23, 31, 32].”
Physically Embedded Planning Problems: New Challenges for Reinforcement Learning (2020, methodology): “, 2018), and progress in this area has largely been driven by the design and engineering of domain-specific hand-crafted and structured methods (Dantam et al., 2016; Kaelbling and Lozano-Pérez, 2011; Plaku and Hager, 2010; Siméon et al., 2004; Srivastava et al., 2014; Toussaint, 2015), which typical”