The final event of the Interreg Sudoe Project COMMANDIA was successfully held last September 30th as an online event. A summary of the project was presented by the coordinator of the project, Youcef Mezouar. After the introduction and summary of the project, the event continued with the presentations from the different partners of the consortium with special emphasis on technical, scientific, dissemination activities and achievements. To conclude, a round table with open live discussion was held with the participants in the event.
You can find a video with that motivates Project COMMANDIA in the following link . Another video here gives an introduction and overview of the project goals. One of the results of the project is a demonstrator that implements some of the techniques developed within the project to manipulate deformable objects. This is explained in this video .
During the COMMANDIA final event we had the opportunity to learn about the work that five partners (SIGMA, U. Alicante, INESCOP, U. Zaragoza, U. Coimbra) developed across three countries (Portugal, Spain, France) on the topic of Collaborative Robotic Mobile Manipulation of Deformable Objects in Industrial Applications. We would also like to take this opportunity to thank all those involved in the success of this project.
Title: Robotic motion coordination based on a geometric deformation measure
Authors: Miguel Aranda, Jose Sanchez, Juan Antonio Corrales Ramon and Youcef Mezouar
Journal: IEEE Systems Journal, doi: 10.1109/JSYST.2021.3107779
Abstract: This article describes a novel approach to achieve motion coordination in a multirobot system based on the concept of deformation. Our main novel contribution is to link these two elements (namely, coordination and deformation). In particular, the core idea of our approach is that the robots’ motions minimize a global measure of the deformation of their positions relative to a prescribed shape. Based on this idea we propose a linear shape controller, that also incorporates a term modeling an affine deformation. We show that the affine term is particularly useful when the deformation to be controlled is large. We also propose controls for the other variables (centroid, rotation, size) that define the geometric configuration of the team. Importantly, these additional controls are completely decoupled from the shape control. The overall approach is simple and robust, and it creates closely coordinated robot motions. Being based on deformation, it is useful in several scenarios involving manipulation tasks: e.g., handling of a highly deformable object, control of an object’s shape, or regulation of the shape formed by the fingertips of a robotic hand. We present simulation and experimental results to validate the proposed approach.
Title: Collision-free Transport of 2D Deformable Objects
Authors: Rafael Herguedas, Gonzalo Lopez-Nicolas, Carlos Sagues
Conference: International Conference on Control, Automation, and Systems (ICCAS 2021), Jeju, Korea, October 12-15, 2021
Abstract: We propose a novel system to transport 2D cloth-like deformable objects with mobile manipulators and without collisions along a known path. First, a new deformation model that allows for real-time shape prediction, based on the paradigm of deformable bounding box, is presented. The transport task is next defined as an optimization problem, which includes a set of linear and nonlinear constraints. These constraints allow to limit the object’s deformations and rotations and to avoid obstacles, respectively. Simulation results are reported to demonstrate the validity of our method.
Title: Multi-scale Laplacian-based FMM for shape control
Authors: Ignacio Cuiral-Zueco and Gonzalo Lopez-Nicolas
Conference: 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). September 27 – October 1, 2021. Prague, Czech Republic
Abstract: Shape control has become a prominent research field as it enables the automation of tasks in many applications. Overall, deforming an object to a desired target shape by using few grippers is a major challenge. The limited information about the object dynamics, the need to combine small and large deformations in order to achieve certain target shapes and the non-linear nature of most deformable objects are factors that significantly hamper shape control performance. In this paper, we propose a shape control method for multi-robot manipulation of large-strain deformable objects. Our approach is based on multi-scale Laplacian descriptors that feed an FMM (Fast Marching Method) for elastic shape contour matching. The FMM’s resulting path and the Laplacian operator are used to define a control strategy for the robot grippers. Simulation experiments carried out with an ARAP (As Rigid As Possible) deformation model provide satisfactory results.
We are happy to announce that INESCOP will present COMMANDIA results in SIMAC 2021.
SIMAC is the International exhibitions of machines and technologies for footwear, leather goods and tanning industry (22-24 September 202, at Fiera Milano Rho, Italy). At SIMAC TANNING TECH you’ll find the most important manufacturers of machines for footwear manufacturing, leather goods and tanneries, accessories and component, chemical products, prototyping systems, lab machinery and equipment, consumables, automation, waste treatment systems and equipment, moulds and die cutters, management and production cycle, logistics, finished-product testing, conveyor systems, etc.
You are welcome to visit us in booth B04 at pavilion 14 from 22nd to 24th of September. See more details here.
You can also take a look to this video for an introduction to the COMMANDIA results.
Different students have presented their final degree projects in the framework of COMMANDIA:
“Development of a practical demonstrator in ROS for the UR10 robot manipulator” by Ignacio Herrera Seara
In this work, the UR10 collaborative robot has been used together with the ROS environment for the development of several practical applications, which have been tested both in simulation and in a real environment. The control of the robot movements has been developed through the MoveIt framework. The implemented applications have been the following: making drawings on a canvas by the robot, extraction and drawing of contours of real objects and, finally, teleoperation of the robot. In addition, the last two applications make use of the RGB-D Realsense D435 camera for their operation.
“Multi-camera mapping with RGB-D sensor” by Juan García-Lechuz Sierra
As with people, many tasks assigned to robots require the coordination of several of them to be carried out. One of them is the perception of the environment, when it has a certain complexity and its analysis is too computationally or time-consuming. For this coordination to be effective, the development of collaborative techniques that take into account the variability of the environment and result in a robust analysis of the environment and as complete as possible is required. The objective of this work is to obtain the representation in three dimensions of different environments, using a moving RGB-D sensor together with an application that allows its development in several sessions, representing the joint work between several cameras placed in different mobile robots to generate a single map. In addition, the My Find Object application is developed to censor objects and surfaces during the generation of the map, processing the images obtained by the RGB-D sensors to eliminate the information of these objects before it is introduced in the application from which the representation in three dimensions is obtained.
“Deformable object manipulation in multi-robot environments” by Andrés Otero García
A case is studied in which multiple IRB120 robots from ABB manipulate simultaneously a deformable object, such as a cloth. The objective is to develop a simulation in which at least two robots can perform synchronized movements to make changes in the deformable object, such as displacement or deformation, without colliding with the object during manipulation.
Title: Enclosing a moving target with an optimally rotated and scaled multiagent pattern
Authors: M. Aranda, Y. Mezouar, G. López-Nicolás, C. Sagüés
Journal: International Journal of Control, vol. 94, no. 3, pp. 601-611, 2021
Abstract: We propose a novel control method to enclose a moving target in a two-dimensional setting with a team of agents forming a prescribed geometric pattern. The approach optimises a measure of the overall agent motion costs, via the minimisation of a suitably defined cost function encapsulating the pattern rotation and scaling. We propose two control laws which use global information and make the agents exponentially converge to the prescribed formation with an optimal scale that remains constant, while the team’s centroid tracks the target. One control law results in a multiagent pattern that keeps a constant orientation in the workspace; for the other, the pattern rotates with constant speed. These behaviors, whose optimality and steadiness are very relevant for the task addressed, occur independently from the target’s velocity. Moreover, the methodology does not require distance measurements, common coordinate references, or communications. We also present formal guarantees of collision avoidance for the proposed approach. Illustrative simulation examples are provided.
The event will first present the framework of COMMANDIA. After the introduction and summary of the project, the event will continue with the presentations from the different partners of the consortium (SIGMA Clermont, INESCOP, Universidad de Zaragoza, Universidad de Alicante, Universidade de Coimbra) with special emphasis on technical, scientific, dissemination activities and achievements. To conclude, a round table with open live discussion will be held with the participants in the event.
Date: 30 September 2021 (Thursday) Time: 9:30 AM to 12:30 PM, Paris time Place: Zoom link
Title: Distributed Linear Control of Multirobot Formations Organized in Triads
Authors: M. Aranda, G. López-Nicolás and Y. Mezouar
Journal: IEEE Robotics and Automation Letters, vol. 6, no. 4, pp. 310-317, Oct. 2021
Abstract: This letter addresses the problem of controlling multiple robots to form a prescribed team shape in two-dimensional space. We consider a team organization in interlaced triads (i.e., groups of three robots). For each triad we define a measure of geometric deformation relative to its prescribed shape. Our main contribution is a novel distributed control law, defined as the gradient descent on the sum of these triangular deformation measures. We show that this geometrically motivated control law is linear, and bears analogies with existing formulations. Moreover, in comparisonwith these formulations our controller is simpler and more flexible to design, converges to the globally optimal shape by construction, and allows analysis of the team size dynamics. We illustrate the proposed approach in simulation.