Advanced dexterous manipulation requires identifying and controlling multiple simultaneous contacts, including interactions with compliant objects where deformations are large. Raw optical tactile images are information rich but lack calibrated physical meaning, limiting cross-sensor use and real-world deployment. We present TensorTouch, a calibration framework that combines finite element analysis and learning to infer dense deformation and stress/force fields from a single tactile image. Across real sensors, TensorTouch achieves contact localization errors under 1.29 mm and mean force errors under 0.139 N per axis. In a multi-object selective grasp task with two simultaneously contacted objects (including identical cables), the system achieves up to 90.0% success. We further demonstrate robustness under repeated loading, yielding 38.295 dB PSNR between the initial tactile image and the image after 20,000 contacts. The learned model runs in real time at 95 Hz on an RTX 5090, proving to be suitable for contact-rich, dexterous manipulation.

SAFER-Splat (Simultaneous Action Filtering and Environment Reconstruction) is a real-time, scalable, and minimally invasive safety filter, based on control barrier functions, for safe robotic navigation in a detailed map constructed at runtime using Gaussian Splatting (GSplat). We propose a novel Control Barrier Function (CBF) that not only induces safety with respect to all Gaussian primitives in the scene, but when synthesized into a controller, is capable of processing hundreds of thousands of Gaussians while maintaining a minimal memory footprint and operating at 15 Hz during online Splat training. Of the total compute time, a small fraction of it consumes GPU resources, enabling uninterrupted training. The safety layer is minimally invasive, correcting robot actions only when they are unsafe. To showcase the safety filter, we also introduce SplatBridge, an open-source software package built with ROS for real-time GSplat mapping for robots. We demonstrate the safety and robustness of our pipeline first in simulation, where our method is 20-50x faster, safer, and less conservative than competing methods based on neural radiance fields. Further, we demonstrate simultaneous GSplat mapping and safety filtering on a drone hardware platform using only on-board perception. We verify that under teleoperation a human pilot cannot invoke a collision. Our videos and codebase can be found at https://chengine.github.io/safer-splat.

We propose a framework for active next best view and touch selection for robotic manipulators using 3D Gaussian Splatting (3DGS). 3DGS is emerging as a useful explicit 3D scene representation for robotics, as it has the ability to represent scenes in a both photorealistic and geometrically accurate manner. However, in real-world, online robotic scenes where the number of views is limited given efficiency requirements, random view selection for 3DGS becomes impractical as views are often overlapping and redundant. We address this issue by proposing an end-to-end online training and active view selection pipeline, which enhances the performance of 3DGS in few-view robotics settings. We first elevate the performance of few-shot 3DGS with a novel semantic depth alignment method using Segment Anything Model 2 (SAM2) that we supplement with Pearson depth and surface normal loss to improve color and depth reconstruction of real-world scenes. We then extend FisherRF, a next-best-view selection method for 3DGS, to select views and touch poses based on depth uncertainty. We perform online view selection on a real robot system during live 3DGS training. We motivate our improvements to few-shot GS scenes, and extend depth-based FisherRF to them, where we demonstrate both qualitative and quantitative improvements on challenging robot scenes. For more information, please see our project page at arm.stanford.edu/next-best-sense.

When working around other agents such as humans, it is important to model their perception capabilities to predict and make sense of their behavior. In this work, we consider agents whose perception capabilities are determined by their limited field of view, viewing range, and the potential to miss objects within their viewing range. By considering the perception capabilities and observation model of agents independently from their motion policy, we show that we can better predict the agents’ behavior; i.e., by reasoning about the perception capabilities of other agents, one can better make sense of their actions. We perform a user study where human operators navigate a cluttered scene while scanning the region for obstacles with a limited field of view and range. We show that by reasoning about the limited observation space of humans, a robot can better learn a human’s strategy for navigate an environment and navigate with minimal collision with dynamic and static obstacles. We also show that this learned model helps it successfully navigate a physical hardware vehicle in real-time. Code available at https://github.com/labicon/HRMotion-RestrictedView.

Manipulation of thin materials is critical for many everyday tasks and remains a significant challenge for robots. While existing research has made strides in tasks like material smoothing and folding, many studies struggle with common failure modes (crumpled corners/edges, incorrect grasp configurations) that a preliminary step of layer detection could solve. We present a novel method for classifying the number of grasped material layers using a custom gripper equipped with DenseTact 2.0 optical tactile sensors. After grasping, the gripper performs an anthropomorphic rubbing motion while collecting optical flow, 6-axis wrench, and joint state data. Using this data in a transformer-based network achieves a test accuracy of 98.21% in classifying the number of grasped cloth layers, and 81.25% accuracy in classifying layers of grasped paper, showing the effectiveness of our dynamic rubbing method. Evaluating different inputs and model architectures highlights the usefulness of tactile sensor information and a transformer model for this task. A comprehensive dataset of 568 labeled trials (368 for cloth and 200 for paper) was collected and made open-source along with this paper.

Dexterous manipulation, especially of small daily objects, continues to pose complex challenges in robotics. This paper introduces the DenseTact-Mini, an optical tactile sensor with a soft, rounded, smooth gel surface and compact design equipped with a synthetic fingernail. We propose three distinct grasping strategies: tap grasping using adhesion forces such as electrostatic and van der Waals, fingernail grasping leveraging rolling/sliding contact between the object and fingernail, and fingertip grasping with two soft fingertips. Through comprehensive evaluations, the DenseTact-Mini demonstrates a lifting success rate exceeding 90.2% when grasping various objects, including items such as 1mm basil seeds, thin paperclips, and items larger than 15mm such as bearings. This work demonstrates the potential of soft optical tactile sensors for dexterous manipulation and grasping.

In this work, we propose a novel method to supervise 3D Gaussian Splatting (3DGS) scenes using optical tactile sensors. Optical tactile sensors have become widespread in their use in robotics for manipulation and object representation; however, raw optical tactile sensor data is unsuitable to directly supervise a 3DGS scene. Our representation leverages a Gaussian Process Implicit Surface to implicitly represent the object, combining many touches into a unified representation with uncertainty. We merge this model with a monocular depth estimation network, which is aligned in a two stage process, coarsely aligning with a depth camera and then finely adjusting to match our touch data. For every training image, our method produces a corresponding fused depth and uncertainty map. Utilizing this additional information, we propose a new loss function, variance-weighted depth supervised loss, for training the 3DGS scene model. We leverage the DenseTact optical tactile sensor and RealSense RGB-D camera to show that combining touch and vision in this manner leads to quantitatively and qualitatively better results than vision or touch alone in few-view scene synthesis on opaque, reflective and transparent objects. Please see our project page at armlabstanford.github.io/touchgs.

Collaborative robots stand to have an immense impact on both human welfare in domestic service applications and industrial superiority in advanced manufacturing with dexterous assembly. The outstanding challenge is providing robotic fingertips with a physical design that makes them adept at performing dexterous tasks that require high-resolution, calibrated shape reconstruction and force sensing. In this work, we present DenseTact 2.0, an optical-tactile sensor capable of visualizing the deformed surface of a soft fingertip and using that image in a neural network to perform both calibrated shape reconstruction and 6-axis wrench estimation. We demon-strate the sensor accuracy of 0.3633mm per pixel for shape reconstruction, 0.410N for forces, 0.387N. mm for torques, and the ability to calibrate new fingers through transfer learning, which achieves comparable performance with only 12% of the non-transfer learning dataset size.

Cooperative table-carrying is a complex task due to the continuous nature of the action and state-spaces, multimodality of strategies, and the need for instantaneous adaptation to other agents. In this work, we present a method for predicting realistic motion plans for cooperative human-robot teams on the task. Using a Variational Recurrent Neural Network (VRNN) to model the variation in the trajectory of a human-robot team across time, we are able to capture the distribution over the team's future states while leveraging information from interaction history. The key to our approach is leveraging human demonstration data to generate trajectories that synergize well with humans during test time in a receding horizon fashion. Comparison between a baseline, sampling-based planner RRT (Rapidly-exploring Random Trees) and the VRNN planner in centralized planning shows that the VRNN generates motion more similar to the distribution of human-human demonstrations than the RRT. Results in a human-in-the-loop user study show that the VRNN planner outperforms decentralized RRT on task-related metrics, and is significantly more likely to be perceived as human than the RRT planner. Finally, we demonstrate the VRNN planner on a real robot paired with a human teleoperating another robot.

Falls are the leading cause of fatal and non-fatal injuries, particularly for older persons. Imbalance can result from the body's internal causes (illness), or external causes (active or passive perturbation). Active perturbation results from applying an external force to a person, while passive perturbation results from human motion interacting with a static obstacle. This work proposes a metric that allows for the monitoring of the persons torso and its correlation to active and passive perturbations. We show that large changes in the torso sway can be strongly correlated to active perturbations. We also show that we can reasonably predict the future path and expected change in torso sway by conditioning the expected path and torso sway on the past trajectory, torso motion, and the surrounding scene. This could have direct future applications to fall prevention. Results demonstrate that the torso sway is strongly correlated with perturbations. And our model is able to make use of the visual cues presented in the panorama and condition the prediction accordingly.

Increasing the performance of tactile sensing in robots enables versatile, in-hand manipulation. Vision-based tactile sensors have been widely used as rich tactile feedback has been shown to be correlated with increased performance in manipulation tasks. Existing tactile sensor solutions with high resolution have limitations that include low accuracy, expensive components, or lack of scalability. In this paper, an inexpensive, scalable, and compact tactile sensor with high-resolution surface deformation modeling for surface reconstruction of the 3D sensor surface is presented. By observing the contact surface with a fisheye camera, it is shown that the surface deformation can be estimated in real-time (1.8 ms) using deep convolutional neural networks. This sensor in its design and sensing abilities represents a significant step toward better object in-hand localization, classification, and surface estimation all enabled by calibrated, high-resolution shape reconstruction.

This paper presents replay overshooting (RO), an algorithm that uses properties of the extended Kalman filter (EKF) to learn nonlinear stochastic latent dynamics models suitable for long-horizon prediction. We build upon overshooting methods used to train other prediction models and recover a novel variational learning objective. Further, we use RO to extend another objective that acts as a surrogate for the true log-likelihood, and show that this objective empirically yields better models than the variational one. We evaluate RO on two tasks: prediction of synthetic video frames of a swinging motorized pendulum and prediction of the planar position of various objects being pushed by a real manipulator (MIT Push Dataset). Our model outperforms several other prediction models on both quantitative and qualitative metrics.

Robotic pouring is an important step in improving the safety, productivity and repeatability in the biotechnology industry and generally increasing the effectiveness of robotics in human based environments. In this work we present a method to autonomously dispense a precise amount of fluid using only visual feedback without using precision pouring instruments such as pipettes, syringes or pourers. We model circular and rectangular pouring container geometries. We prove that for square containers we can control the flow by only observing the fluid height in the receiving beaker. We show a systematic approach using a hybrid control scheme that is robust to the initial amount of fluid in the pouring container and inconsistent flow. Specifically we present (a) a model for pouring (b) a model based algorithm to drive a robot arm (c) visual feedback for regulating the pouring rate. We demonstrate this using the Rethink Robotics Sawyer manipulator and mvBluefox MLC202bc camera.

We present a new descriptor for tactile 3D object classification. It is invariant to object movement and simple to construct, using only the relative geometry of points on the object surface. We demonstrate successful classification of 185 objects in 10 categories, at sparse to dense surface sampling rate in point cloud simulation, with an accuracy of 77.5% at the sparsest and 90.1% at the densest. In a physics-based simulation, we show that contact clouds resembling the object shape can be obtained by a series of gripper closures using a robotic hand equipped with sparse tactile arrays. Despite sparser sampling of the object's surface, classification still performs well, at 74.7%. On a real robot, we show the ability of the descriptor to discriminate among different object instances, using data collected by a tactile hand.

Transport of individual cells or chemical payloads on a subcellular scale is an enabling tool for the study of cellular communication, cell migration, and other localized phenomena. We present a magnetically actuated robotic system for the fully automated manipulation of cells and microbeads. Our strategy uses autofluorescent robotic transporters and fluorescently labeled microbeads to aid tracking and control in optically obstructed environments. We demonstrate automated delivery of microbeads infused with chemicals to specified positions on neurons.