SAFE: Unified Slip and Fracture Detection with Low-Cost Acoustic Sensing in Robotic Grasping

Summary

Abstract Summary

Manipulating fragile objects remains challenging as robots must understand the state of what they grasp, such as slip or fracture, to respond appropriately, especially when material properties are unknown. In this paper, we present SAFE: a low-cost, general-purpose sensing approach that detects both slip and fracture in real time using two passive polyvinylidene fluoride (PVDF) acoustic sensors and motor proprioception, without relying on vision or prior material knowledge. The sensors are mounted on a compliant Fin Ray gripper, and a unified HistGradientBoosting classifier reports the state (normal, slip, or fracture) from a 79-dimensional feature vector. Under leave-one-grasp-out cross-validation, SAFE achieves an Alert-F1 of 0.884 with near-zero slip-fracture confusion, and ablations confirm that acoustic sensing is indispensable. An adaptive grasp controller built on this detection layer runs at 104 Hz on a Jetson Orin Nano, achieving 91.3% success across 46 closed-loop robot trials spanning diverse object categories, while each fixed-force strategy drops to 0% on object conditions that mismatch its preset. It further reaches 82.4% success on novel objects unseen during training, demonstrating robust, failure-aware grasp control without object-specific calibration.

Key Points

Manipulating fragile objects remains challenging as robots must understand the state of what they grasp, such as slip or fracture, to respond appropriately, especially when material properties are unknown.
In this paper, we present SAFE: a low-cost, general-purpose sensing approach that detects both slip and fracture in real time using two passive polyvinylidene fluoride (PVDF) acoustic sensors and motor proprioception, without relying on vision or prior material knowledge.
The sensors are mounted on a compliant Fin Ray gripper, and a unified HistGradientBoosting classifier reports the state (normal, slip, or fracture) from a 79-dimensional feature vector.
Under leave-one-grasp-out cross-validation, SAFE achieves an Alert-F1 of 0.884 with near-zero slip-fracture confusion, and ablations confirm that acoustic sensing is indispensable.
An adaptive grasp controller built on this detection layer runs at 104 Hz on a Jetson Orin Nano, achieving 91.3% success across 46 closed-loop robot trials spanning diverse object categories, while each fixed-force strategy drops to 0% on object conditions that mismatch its preset.

Methodology

Real-time detection of slip and fracture during grasping must satisfy three constraints: it cannot rely on prior knowledge of object material, it must run within a real-time control loop on low-cost edge hardware, and it must resolve two physically different failure signatures from a single, minimal sensing layer.

We address these with compliant hardware carrying acoustic/proprioceptive sensing, a windowed feature representation, and a unified classifier.

Methodology — SAFE: Unified Slip and Fracture Detection with Low-Cost Acoustic Sensing in Robotic Grasping — Figure 0: Objects used for data collection and evaluation. Training set: two fracture-risk objects, eggshell (a) and potato chips (b), and three slip-risk objects with varied weights, hand-soap bottle (c), paper cup (d), and water bottle (e). Testing set (out-of-distribution): onion ring chips (f), glass beaker coated with soap (g), and plastic cup (h); the latter two were collected with varied weights.

Experiments

Table 2 reports classification performance under the Leave-One-Grasp-Out protocol described in Section III-D . The proposed model achieved an Alert-F1 (the mean of slip and fracture F1: the normal class comprises 71% of samples and would otherwise dominate an aggregate score) of 0.884 [0.856, 0.908], with slip and fracture individually reaching F1 of 0.916 and 0.851.

The 95% confidence interval (CI) was estimated using a group-level cluster bootstrap with 5000 resamples over grasp-trial groups. As shown in the confusion matrix (Fig.

Experiments — SAFE: Unified Slip and Fracture Detection with Low-Cost Acoustic Sensing in Robotic Grasping — Figure 1: Median magnitude spectrum for each class on the outer channel, across 1499 windows, showing visible class separability directly in the frequency domain.

Sources and Demonstrations

Figures are reproduced from Wang et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.