FlatLab: A Unified Methodology Framework and Simulation-Based Benchmark for Robotic Manipulation of Flat Objects

Xingyu Zhu, Wenshuo Han, Zhouyu Wang, Yuran Wang, Ruihai Wu, Hao Dong, Fan Tang, Hechang Chen, Hyung Jin Chang, Yixing Gao

TL;DR: A unified strategy-action decoupling framework enables generalizable robotic manipulation of diverse flat objects in FlatLab, a new simulation benchmark.

Abstract: Robotic manipulation of flat objects is challenging due to the ungraspable configurations and strong variations in object geometry and material. Existing methods rely on heuristic pre-manipulation and are often evaluated in closed settings with limited generalization. We propose a unified framework that decouples the manipulation into a strategy generator and an action execution module. The strategy generator predicts appropriate manipulation strategies from object point clouds by learning strategy-centric, object-invariant representations via simulated data transformation and contrastive learning. Conditioned on the predicted strategy, the execution module decomposes long-horizon manipulation into reusable action primitives and dynamically composes them to generate stable trajectories. To enable systematic evaluation, we introduce FlatLab, a comprehensive simulation benchmark for robotic flat object manipulation. FlatLab provides high-fidelity physical simulation of diverse rigid and deformable flat objects, automated multi-modal data collection, and standardized task definitions and evaluation protocols. Experiments conducted in FlatLab demonstrate that our approach generalizes effectively to unseen objects and categories, outperforming existing baselines. The project page and the code are provided at \url{https://flatlab-web.github.io/}.

Lay Summary: This paper presents FlatLab, a robotic simulation platform for manipulating flat objects such as books, keyboards, and fabrics. Flat objects are difficult for robots to grasp because they often lie flush against surfaces and provide few accessible grasp positions. Existing methods usually rely on a single fixed strategy and struggle to generalize across different object shapes and materials. Our method enables robots to adaptively select different manipulation strategies, such as sliding thin objects to a table edge, lifting larger rigid objects with two arms, or squeezing deformable objects to create graspable folds. By separating strategy selection from action execution, the framework achieves more generalizable manipulation. FlatLab is a large-scale simulation benchmark with over 100 flat objects, realistic physics simulation, automated data collection, and standardized evaluation tasks. Experiments in both simulation and real-world settings demonstrate strong performance and generalization to unseen objects and categories.

Link To Code: https://flatlab-web.github.io/

Primary Area: Applications->Robotics

Keywords: Robotic Manipulation, Simulation Based Benchmark, Flat Objects Grasping

Submission Number: 15924