Robotics (cs.RO)

Abstract: Large Language models (LLMs) have shown remarkable success in assisting robot learning tasks, i.e., complex household planning. However, the performance of pretrained LLMs heavily relies on domain-specific templated text data, which may be infeasible in real-world robot learning tasks with image-based observations. Moreover, existing LLMs with text inputs lack the capability to evolve with non-expert interactions with environments. In this work, we introduce a novel learning paradigm that generates robots' executable actions in the form of text, derived solely from visual observations, using language-based summarization of these observations as the connecting bridge between both domains. Our proposed paradigm stands apart from previous works, which utilized either language instructions or a combination of language and visual data as inputs. Moreover, our method does not require oracle text summarization of the scene, eliminating the need for human involvement in the learning loop, which makes it more practical for real-world robot learning tasks. Our proposed paradigm consists of two modules: the SUM module, which interprets the environment using visual observations and produces a text summary of the scene, and the APM module, which generates executable action policies based on the natural language descriptions provided by the SUM module. We demonstrate that our proposed method can employ two fine-tuning strategies, including imitation learning and reinforcement learning approaches, to adapt to the target test tasks effectively. We conduct extensive experiments involving various SUM/APM model selections, environments, and tasks across 7 house layouts in the VirtualHome environment. Our experimental results demonstrate that our method surpasses existing baselines, confirming the effectiveness of this novel learning paradigm.

Abstract: Improving the generalization capabilities of general-purpose robotic agents has long been a significant challenge actively pursued by research communities. Existing approaches often rely on collecting large-scale real-world robotic data, such as the RT-1 dataset. However, these approaches typically suffer from low efficiency, limiting their capability in open-domain scenarios with new objects, and diverse backgrounds. In this paper, we propose a novel paradigm that effectively leverages language-grounded segmentation masks generated by state-of-the-art foundation models, to address a wide range of pick-and-place robot manipulation tasks in everyday scenarios. By integrating precise semantics and geometries conveyed from masks into our multi-view policy model, our approach can perceive accurate object poses and enable sample-efficient learning. Besides, such design facilitates effective generalization for grasping new objects with similar shapes observed during training. Our approach consists of two distinct steps. First, we introduce a series of foundation models to accurately ground natural language demands across multiple tasks. Second, we develop a Multi-modal Multi-view Policy Model that incorporates inputs such as RGB images, semantic masks, and robot proprioception states to jointly predict precise and executable robot actions. Extensive real-world experiments conducted on a Franka Emika robot arm validate the effectiveness of our proposed paradigm. Real-world demos are shown in YouTube (https://www.youtube.com/watch?v=1m9wNzfp_4E ) and Bilibili (https://www.bilibili.com/video/BV178411Z7H2/ ).

Abstract: High-fidelity datasets play a pivotal role in imbuing simulators with realism, enabling the benchmarking of various state-of-the-art deep inference models. These models are particularly instrumental in tasks such as semantic segmentation, classification, and localization. This study showcases the efficacy of a customized manufacturing dataset comprising 60 classes in the creation of a high-fidelity digital twin of a robotic manipulation environment. By leveraging the concept of transfer learning, different 6D pose estimation models are trained within the simulated environment using domain randomization and subsequently tested on real-world objects to assess domain adaptation. To ascertain the effectiveness and realism of the created data-set, pose accuracy and mean absolute error (MAE) metrics are reported to quantify the model2real gap.

Abstract: Endowing robots with tactile capabilities opens up new possibilities for their interaction with the environment, including the ability to handle fragile and/or soft objects. In this work, we equip the robot gripper with low-cost vision-based tactile sensors and propose a manipulation algorithm that adapts to both rigid and soft objects without requiring any knowledge of their properties. The algorithm relies on a touch and slip detection method, which considers the variation in the tactile images with respect to reference ones. We validate the approach on seven different objects, with different properties in terms of rigidity and fragility, to perform unplugging and lifting tasks. Furthermore, to enhance applicability, we combine the manipulation algorithm with a grasp sampler for the task of finding and picking a grape from a bunch without damaging~it.

Abstract: Usage of purely vision based solutions for row switching is not well explored in existing vision based crop row navigation frameworks. This method only uses RGB images for local feature matching based visual feedback to exit crop row. Depth images were used at crop row end to estimate the navigation distance within headland. The algorithm was tested on diverse headland areas with soil and vegetation. The proposed method could reach the end of the crop row and then navigate into the headland completely leaving behind the crop row with an error margin of 50 cm.

Abstract: The development of tactile sensing and its fusion with computer vision is expected to enhance robotic systems in handling complex tasks like deformable object manipulation. However, readily available industrial grippers typically lack tactile feedback, which has led researchers to develop and integrate their own tactile sensors. This has resulted in a wide range of sensor hardware, making it difficult to compare performance between different systems. We highlight the value of accessible open-source sensors and present a set of fingertips specifically designed for fine object manipulation, with readily interpretable data outputs. The fingertips are validated through two difficult tasks: cloth edge tracing and cable tracing. Videos of these demonstrations, as well as design files and readout code can be found at https://github.com/RemkoPr/icra-2023-workshop-tactile-fingertips.