1. |
- Almeida, Diogo, 1991-, et al.
(author)
-
Bimanual folding assembly: Switched control and contact point estimation
- 2016
-
In: 16th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2016; Hotel WestinCancun; Mexico; 15-17 November 2016. - Cancun : IEEE. - 2164-0580. - 9781509047185 ; , s. Art no 7803279, Pages 210-216
-
Conference paper (peer-reviewed)abstract
- Robotic assembly in unstructured environments is a challenging task, due to the added uncertainties. These can be mitigated through the employment of assembly systems, which offer a modular approach to the assembly problem via the conjunction of primitives. In this paper, we use a dual-arm manipulator in order to execute a folding assembly primitive. When executing a folding primitive, two parts are brought into rigid contact and posteriorly translated and rotated. A switched controller is employed in order to ensure that the relative motion of the parts follows the desired model, while regulating the contact forces. The control is complemented with an estimator based on a Kalman filter, which tracks the contact point between parts based on force and torque measurements. Experimental results are provided, and the effectiveness of the control and contact point estimation is shown.
|
|
2. |
- Bergonzani, Ivan, et al.
(author)
-
Fast Dynamic Walking with RH5 Humanoid Robot
- 2023
-
In: IEEE-RAS International Conference on Humanoid Robots. - 2164-0572 .- 2164-0580.
-
Conference paper (peer-reviewed)abstract
- Humanoid robots have the potential of becoming general purpose robots augmenting the human work-force in industries. However, they must match the agility and versatility of humans. It is particularly challenging for humanoids actuated with electric drives to achieve that as one must strive for the right balance between mass-inertial distribution in the robot as well as velocity and force transmissions in its actuation concept. In addition to optimal design of the robot, the control system must be designed to exploit the full potential of the robot. In this paper, we perform experimental investigations on the dynamic walking capabilities of a series-parallel hybrid humanoid named RH5. We demonstrate that it is possible to walk up to speeds of 0.43 m/s with a position controlled robot without full state feedback which makes it one of the fastest walking humanoids with similar size and actuation modalities. Video of the experiments is available at: https://youtu.be/39GL2vPedGY
|
|
3. |
- Björkman, Mårten, et al.
(author)
-
Learning to Disambiguate Object Hypotheses through Self-Exploration
- 2014
-
In: 14th IEEE-RAS International Conference onHumanoid Robots. - : IEEE Computer Society. - 2164-0572 .- 2164-0580. - 9781479971749 - 9781479971756
-
Conference paper (peer-reviewed)abstract
- We present a probabilistic learning framework to form object hypotheses through interaction with the environment. A robot learns how to manipulate objects through pushing actions to identify how many objects are present in the scene. We use a segmentation system that initializes object hypotheses based on RGBD data and adopt a reinforcement approach to learn the relations between pushing actions and their effects on object segmentations. Trained models are used to generate actions that result in minimum number of pushes on object groups, until either object separation events are observed or it is ensured that there is only one object acted on. We provide baseline experiments that show that a policy based on reinforcement learning for action selection results in fewer pushes, than if pushing actions were selected randomly.
|
|
4. |
|
|
5. |
- Viña, Francisco, 1990-, et al.
(author)
-
Predicting Slippage and Learning Manipulation Affordances through Gaussian Process Regression
- 2013
-
In: IEEE-RAS International Conference on Humanoid Robots. - : IEEE Computer Society. - 2164-0580 .- 2164-0572. ; , s. 462-468
-
Conference paper (peer-reviewed)abstract
- Object grasping is commonly followed by some form of object manipulation - either when using the grasped object as a tool or actively changing its position in the hand through in-hand manipulation to afford further interaction. In this process, slippage may occur due to inappropriate contact forces, various types of noise and/or due to the unexpected interaction or collision with the environment. In this paper, we study the problem of identifying continuous bounds on the forces and torques that can be applied on a grasped object before slippage occurs. We model the problem as kinesthetic rather than cutaneous learning given that the measurements originate from a wrist mounted force-torque sensor. Given the continuous output, this regression problem is solved using a Gaussian Process approach. We demonstrate a dual armed humanoid robot that can autonomously learn force and torque bounds and use these to execute actions on objects such as sliding and pushing. We show that the model can be used not only for the detection of maximum allowable forces and torques but also for potentially identifying what types of tasks, denoted as manipulation affordances, a specific grasp configuration allows. The latter can then be used to either avoid specific motions or as a simple step of achieving in-hand manipulation of objects through interaction with the environment.
|
|
6. |
- Ambrus, Rares, et al.
(author)
-
Unsupervised object segmentation through change detection in a long term autonomy scenario
- 2016
-
In: IEEE-RAS International Conference on Humanoid Robots. - : IEEE. - 9781509047185 ; , s. 1181-1187
-
Conference paper (peer-reviewed)abstract
- In this work we address the problem of dynamic object segmentation in office environments. We make no prior assumptions on what is dynamic and static, and our reasoning is based on change detection between sparse and non-uniform observations of the scene. We model the static part of the environment, and we focus on improving the accuracy and quality of the segmented dynamic objects over long periods of time. We address the issue of adapting the static structure over time and incorporating new elements, for which we train and use a classifier whose output gives an indication of the dynamic nature of the segmented elements. We show that the proposed algorithms improve the accuracy and the rate of detection of dynamic objects by comparing with a labelled dataset.
|
|
7. |
- Caccamo, Sergio, et al.
(author)
-
Active perception and modeling of deformable surfaces using Gaussian processes and position-based dynamics
- 2016
-
In: IEEE-RAS International Conference on Humanoid Robots. - : IEEE. - 9781509047185 ; , s. 530-537
-
Conference paper (peer-reviewed)abstract
- Exploring and modeling heterogeneous elastic surfaces requires multiple interactions with the environment and a complex selection of physical material parameters. The most common approaches model deformable properties from sets of offline observations using computationally expensive force-based simulators. In this work we present an online probabilistic framework for autonomous estimation of a deformability distribution map of heterogeneous elastic surfaces from few physical interactions. The method takes advantage of Gaussian Processes for constructing a model of the environment geometry surrounding a robot. A fast Position-based Dynamics simulator uses focused environmental observations in order to model the elastic behavior of portions of the environment. Gaussian Process Regression maps the local deformability on the whole environment in order to generate a deformability distribution map. We show experimental results using a PrimeSense camera, a Kinova Jaco2 robotic arm and an Optoforce sensor on different deformable surfaces.
|
|