Selected work

Representative Papers

The Rosario Dataset v2: Multimodal Dataset for Agricultural Robotics

Aug 29, 2025

Agricultural robots face significant challenges in localization, mapping, and navigation under natural illumination variations, motion blur, uneven terrain, and long-range visual aliasing—exacerbated by the absence of high-synchronization, ground-truth–annotated multimodal benchmark datasets. To address this, we present and publicly release the first high-precision, multimodal SLAM dataset specifically designed for soybean field environments. It integrates synchronized stereo infrared/RGB cameras, IMU, multi-mode GNSS, and wheel odometry, with hardware-level timestamp synchronization and post-processed differential GNSS to deliver centimeter-accurate 6-DOF ground-truth trajectories and long-distance loop closures. The dataset comprises over two hours of real-world field sequences. We systematically evaluate state-of-the-art multimodal SLAM methods, identifying critical performance bottlenecks. This work fills a key gap in agricultural SLAM evaluation, enabling reproducible algorithm development and standardized benchmarking.

0 citationsRead paper

Encoding and Reasoning About Arrays in Set Theory

Aug 15, 2025

Existing set-theoretic automated reasoning tools—such as {log}—lack native support for arrays, limiting their applicability to programs involving mixed data structures. Method: We propose encoding arrays as functions represented by sets of ordered pairs, thereby reducing array reasoning to pure set-theoretic reasoning. To formalize this, we define a decidable fragment of set theory extended with function and array semantics, and implement its solver using constraint logic programming. Contribution/Results: This work introduces the first unified formalization and automated reasoning framework for arrays, sets, and relations within {log}, overcoming the prior decidability barrier posed by array constructs. Experimental evaluation demonstrates that our approach effectively encodes and solves programs featuring nontrivial array operations—including indexing, update, and length constraints—thereby substantially enhancing {log}’s capability to reason about heterogeneous data structures.

0 citationsRead paper

Optimizing Exploration with a New Uncertainty Framework for Active SLAM Systems

Jun 21, 2025

To address the exploration-exploitation imbalance, trajectory-dependent mapping uncertainty, and lack of general stopping criteria in active SLAM, this paper proposes an active mapping framework grounded in Uncertainty Maps (UM) and Uncertainty Frontiers (UF). We introduce Signed Relative Entropy (SiREn), a novel metric that jointly quantifies spatial coverage and state uncertainty, enabling dynamic balancing between exploration and exploitation. A probabilistic UM model is developed to support heterogeneous sensors—including monocular/stereo cameras, LiDAR, and multi-sensor fusion. Furthermore, we design a UF-driven online planning algorithm coupled with an adaptive termination strategy. To our knowledge, this is the first approach achieving fully autonomous exploration in open environments, significantly improving both mapping accuracy and path efficiency. We release an open-source ROS implementation alongside benchmark real-world and simulated datasets, facilitating reproducible research and community extension.

0 citationsRead paper

From a Constraint Logic Programming Language to a Formal Verification Tool

May 23, 2025

This paper addresses the semantic gap between programming and specification in formal verification by extending the {log} constraint logic programming language into an integrated verification framework that unifies program execution and automated proof. Methodologically, it constructs an executable state-machine model grounded in set theory and binary relations, enabling unified support for modeling, scenario execution, verification condition generation, SMT-based automated proving, and test-case generation. Crucially, it achieves, for the first time, dual semantics—where the same set-theoretic code serves both as an executable program and a formal specification. Contributions include: (1) eliminating the semantic divide between programming and verification; and (2) establishing an end-to-end verification environment that achieves fully automated security verification and high-coverage test generation on multiple industrial-scale protocols, with verification efficiency substantially surpassing traditional approaches.

0 citationsRead paper

Recent publications

Latest Papers

The Rosario Dataset v2: Multimodal Dataset for Agricultural Robotics

Aug 29, 2025

Agricultural robots face significant challenges in localization, mapping, and navigation under natural illumination variations, motion blur, uneven terrain, and long-range visual aliasing—exacerbated by the absence of high-synchronization, ground-truth–annotated multimodal benchmark datasets. To address this, we present and publicly release the first high-precision, multimodal SLAM dataset specifically designed for soybean field environments. It integrates synchronized stereo infrared/RGB cameras, IMU, multi-mode GNSS, and wheel odometry, with hardware-level timestamp synchronization and post-processed differential GNSS to deliver centimeter-accurate 6-DOF ground-truth trajectories and long-distance loop closures. The dataset comprises over two hours of real-world field sequences. We systematically evaluate state-of-the-art multimodal SLAM methods, identifying critical performance bottlenecks. This work fills a key gap in agricultural SLAM evaluation, enabling reproducible algorithm development and standardized benchmarking.

0 citationsRead paper

Encoding and Reasoning About Arrays in Set Theory

Aug 15, 2025

Existing set-theoretic automated reasoning tools—such as {log}—lack native support for arrays, limiting their applicability to programs involving mixed data structures. Method: We propose encoding arrays as functions represented by sets of ordered pairs, thereby reducing array reasoning to pure set-theoretic reasoning. To formalize this, we define a decidable fragment of set theory extended with function and array semantics, and implement its solver using constraint logic programming. Contribution/Results: This work introduces the first unified formalization and automated reasoning framework for arrays, sets, and relations within {log}, overcoming the prior decidability barrier posed by array constructs. Experimental evaluation demonstrates that our approach effectively encodes and solves programs featuring nontrivial array operations—including indexing, update, and length constraints—thereby substantially enhancing {log}’s capability to reason about heterogeneous data structures.

0 citationsRead paper

Optimizing Exploration with a New Uncertainty Framework for Active SLAM Systems

Jun 21, 2025

To address the exploration-exploitation imbalance, trajectory-dependent mapping uncertainty, and lack of general stopping criteria in active SLAM, this paper proposes an active mapping framework grounded in Uncertainty Maps (UM) and Uncertainty Frontiers (UF). We introduce Signed Relative Entropy (SiREn), a novel metric that jointly quantifies spatial coverage and state uncertainty, enabling dynamic balancing between exploration and exploitation. A probabilistic UM model is developed to support heterogeneous sensors—including monocular/stereo cameras, LiDAR, and multi-sensor fusion. Furthermore, we design a UF-driven online planning algorithm coupled with an adaptive termination strategy. To our knowledge, this is the first approach achieving fully autonomous exploration in open environments, significantly improving both mapping accuracy and path efficiency. We release an open-source ROS implementation alongside benchmark real-world and simulated datasets, facilitating reproducible research and community extension.

0 citationsRead paper

From a Constraint Logic Programming Language to a Formal Verification Tool

May 23, 2025

This paper addresses the semantic gap between programming and specification in formal verification by extending the {log} constraint logic programming language into an integrated verification framework that unifies program execution and automated proof. Methodologically, it constructs an executable state-machine model grounded in set theory and binary relations, enabling unified support for modeling, scenario execution, verification condition generation, SMT-based automated proving, and test-case generation. Crucially, it achieves, for the first time, dual semantics—where the same set-theoretic code serves both as an executable program and a formal specification. Contributions include: (1) eliminating the semantic divide between programming and verification; and (2) establishing an end-to-end verification environment that achieves fully automated security verification and high-coverage test generation on multiple industrial-scale protocols, with verification efficiency substantially surpassing traditional approaches.

0 citationsRead paper