An official website of the United States government
Detecting and localizing contacts is essential for robot manipulators to perform contact-rich tasks in unstructured environments. While robot skins can localize contacts on the surface of robot arms, these sensors are not yet robust or easily accessible. As such, prior works have explored using proprioceptive observations, such as joint velocities and torques, to perform contact localization. Many past approaches assume the robot is static during contact incident, a single contact is made at a time, or having access to accurate dynamics models and joint torque sensing. In this work, we relax these assumptions and propose using Domain Randomization to train a neural network to localize contacts of robot arms in motion without joint torque observations. Our method uses a novel cylindrical projection encoding of the robot arm surface, which allows the network to use convolution layers to process input features and transposed convolution layers to predict contacts. The trained network achieves a contact detection accuracy of 91.5% and a mean contact localization error of 3.0cm. We further demonstrate an application of the contact localization model in an obstacle mapping task, evaluated in both simulation and the real world.
more »
« less
Contact Localization via Active Oscillatory Actuation
Many robotic tasks rely on physical interactions with the task environment. Sensing when and where links make physical contacts can be crucial in several applications, including but not limited to grasping, locomotion, collaborative robotics and navigation. While sensorizing robot end effectors with intrinsic tactile devices is a logical approach, current and accessible options are often expensive or require invasive modifications. This paper presents a prototype method of both sensing and localizing contacts along a rigid link that can be readily incorporated into existing machines. The mechanism is lightweight and low-cost, and functions by actively providing an oscillatory mechanical actuation signal to a rigid link, whose mechanical response is measured with an inertial device and is used to localize touch at one of five designated contact points. Classification is performed with supervised methods using transient behavior and spectral features. Evaluation is conducted with five-fold cross validation, and preliminary results indicate promising performance in localizing the point of contact on the rigid link with accuracy of over 97%.
more »
« less
- Award ID(s):
- 2101107
- PAR ID:
- 10326945
- Date Published:
- Journal Name:
- 2022 IEEE/SICE International Symposium on System Integration (SII)
- Page Range / eLocation ID:
- 344 to 350
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Tan, Jie; Toussaint, Marc; Darvish, Kourosh (Ed.)Contacts play a critical role in most manipulation tasks. Robots today mainly use proximal touch/force sensors to sense contacts, but the information they provide must be calibrated and is inherently local, with practical applications relying either on extensive surface coverage or restrictive assumptions to resolve ambiguities. We propose a vision-based extrinsic contact localization task: with only a single RGB-D camera view of a robot workspace, identify when and where an object held by the robot contacts the rest of the environment. We show that careful task-attuned design is critical for a neural network trained in simulation to discover solutions that transfer well to a real robot. Our final approach im2contact demonstrates the promise of versatile general-purpose contact perception from vision alone, performing well for localizing various contact types (point, line, or planar; sticking, sliding, or rolling; single or multiple), and even under occlusions in its camera view. Video results can be found at: https://sites.google.com/view/im2contact/homemore » « less
-
null (Ed.)Contacts between robots and environment are often assumed to be rigid for control purposes. This assumption can lead to poor performance when contacts are soft and/or underdamped. However, the problem of balancing on soft contacts has not received much attention in the literature. This paper presents two novel approaches to control a legged robot balancing on visco-elastic contacts, and compares them to other two state-of-the-art methods. Our simulation results show that performance heavily depends on the contact stiffness and the noises/uncertainties introduced in the simulation. Briefly, the two novel controllers performed best for soft/medium contacts, whereas “inverse-dynamics control under rigid-contact assumptions” was the best one for stiff contacts. Admittance control was instead the most robust, but suffered in terms of performance. These results shed light on this challenging problem, while pointing out interesting directions for future investigation.more » « less
-
We characterize the mechanical recovery of compliant silicone gels following adhesive contact failure. We establish broad, stable adhesive contacts between rigid microspheres and soft gels, then stretch the gels to large deformations by pulling quasi-statically on the contact. Eventually, the adhesive contact begins to fail, and ultimately slides to a final contact point on the bottom of the sphere. Immediately after detachment, the gel recoils quickly with a self-similar surface profile that evolves as a power law in time, suggesting that the adhesive detachment point is singular. The singular dynamics we observe are consistent with a relaxation process driven by surface stress and slowed by viscous flow through the porous, elastic network of the gel. Our results emphasize the importance of accounting for both the liquid and solid phases of gels in understanding their mechanicsmore » « less
-
Abstract Background Individual behavioural decisions are responses to a person’s perceived social norms that could be shaped by both their physical and social environment. In the context of the COVID-19 pandemic, these environments correspond to epidemiological risk from contacts and the social construction of risk by communication within networks of friends. Understanding the circumstances under which the influence of these different social networks can promote the acceptance of non-pharmaceutical interventions and consequently the adoption of protective behaviours is critical for guiding useful, practical public health messaging. Methods We explore how information from both physical contact and social communication layers of a multiplex network can contribute to flattening the epidemic curve in a community. Connections in the physical contact layer represent opportunities for transmission, while connections in the communication layer represent social interactions through which individuals may gain information, e.g. messaging friends. Results We show that maintaining focus on awareness of risk among each individual’s physical contacts promotes the greatest reduction in disease spread, but only when an individual is aware of the symptoms of a non-trivial proportion of their physical contacts (~ ≥ 20%). Information from the social communication layer without was less useful when these connections matched less well with physical contacts and contributed little in combination with accurate information from physical contacts. Conclusions We conclude that maintaining social focus on local outbreak status will allow individuals to structure their perceived social norms appropriately and respond more rapidly when risk increases. Finding ways to relay accurate local information from trusted community leaders could improve mitigation even where more intrusive/costly strategies, such as contact-tracing, are not possible.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/SII52469.2022.9708753
