Fabio Dalla Libera

Adaptive foraging for simulated and real robotic swarms: the dynamical response threshold approach

Developing self-organised swarm systems capable of adapting to environmental changes as well as to dynamic situations is a complex challenge. An efficient labour division model, with the ability to regulate the distribution of work among swarm robots, is an important element of this kind of system. This paper extends the popular response threshold model and proposes a new adaptive response threshold model (ARTM). Experiments were carried out in simulation and in real-robot scenarios with the aim of studying the performance of this new adaptive model. Results presented in this paper verify that the extended approach improves on the adaptability of previous systems. For example, by reducing collision duration among robots in foraging missions, our approach helps small swarms of robots to adapt more efficiently to changing environments, thus increasing their self-sustainability (survival rate). Finally, we propose a minimal version of ARTM, which is derived from the conclusions drawn through real-robot and simulation results.

Teaching by touching: An intuitive method for development of humanoid robot motions

This paper investigates touching as a natural way for humans to communicate with robots. In particular we developed a system to edit motions of a small humanoid robot by touching its body parts. This interface has two purposes: it allows the user to develop robot motions in a very intuitive way, and it allows us to collect data useful for studying the characteristics of touching as a means of communication. Experimental results confirm the interfaces ease of use for inexpert users, and analysis of the data collected during human-robot teaching episodes has yielded several useful insights.

Direct programming of a central pattern generator for periodic motions by touching

Much of the literature shows that Central Pattern Generators (CPGs) are a good approach for generating periodic motions for legged robots. In most of the presented works the numerous CPG parameters are set by automatic techniques like genetic algorithms. This gives the user little control over the resulting motions, since all of the desired features of the motion must be encoded by a fitness/score function. In this paper we present the idea of setting the CPG parameters by interaction with the user, in particular by using the tactile interaction. Key elements of the system are a CPG network, a touch protocol and a self-collision prevention system. In this paper we present a practical implementation of each element that confirms the feasibility of the method.

Developing Robot Motions by Simulated Touch Sensors

Touch is a very powerful but not much studied communication mean in human-robot interaction. Nonetheless many robots are not equipped with touch sensors, because it is often difficult to place such sensors over the robot surface or simply because the main task of the robot does not require them. We propose an approach that allows developing motions for a real humanoid robot by touching its 3D representation. This simulated counterpart can be equipped with touch sensors not physically available and allows the user to interact with a robot moving in slow-play, which is not possible in real world due to the changes in the dynamics. The developed interface, employing simulated touch sensors, allows inexperienced users to program robot movements in an intuitive way without any modification of the robots hardware. Thanks to this tool we can also study how humans employ touch for communication. We then report how simulation can be used to study user dependence of touch instructions assuring all the subjects to be in exactly the same conditions.

Stochastic resonance induced continuous activation functions in a neural network consisting of threshold elements

Stochastic resonance (SR) is a phenomenon occurring in some nonlinear systems by which a signal provided as input that is too small in magnitude to normally influence the systems output can actually influence the systems output once a non-zero level of noise is provided. SR has been extensively studied both theoretically and experimentally, and noise has been exploited for improving the performance in both biological and artificial systems. In addition to its scientific importance, the use of noise has attracted interest because of its potential to overcome the present limitations in engineering applications. In this study, we investigate the use of a universal approximator exploiting SR as a new realization of well-established feed-forward neural networks. The proposed universal approximator consists of groups of threshold elements. Although the approximation universality of a network consisting of threshold elements has been proven in terms of extreme learning machine implementations, once SR is taken into account, the system can be modeled in a form identical to that of a classic generic three-layered neural network, for which the universal approximation capability has been proven. The capability of the proposed approximator for serving as a universal approximator is first proven theoretically in the limit of an infinite number of hidden units. Subsequently, the performance achieved by the backpropagation type and the extreme learning machine type learning algorithms is experimentally evaluated for cases involving limited numbers of hidden units, highlighting the SR effect occurring in the proposed system.

Clustering of Humanoid Robot Motions Executed in Response to Touch

We perform a study on responses that should be performed by robots when touched by humans. To study the kinds of robot responses in general, there is a need for clustering similar responses. We present the use of Multiple correspondence analysis (MCA) and hierarchical clustering method as a way of clustering different humanoid robot postures. MCA is commonly used to analyze data with discrete variables. Since clustering solely by MCA was impractical for our application, hierarchical clustering method was performed to aid the direct inspection of the robot response clusters.

Extracting motor synergies from random movements for low-dimensional task-space control of musculoskeletal robots.

In the field of human motor control, the motor synergy hypothesis explains how humans simplify body control dimensionality by coordinating groups of muscles, called motor synergies, instead of controlling muscles independently. In most applications of motor synergies to low-dimensional control in robotics, motor synergies are extracted from given optimal control signals. In this paper, we address the problems of how to extract motor synergies without optimal data given, and how to apply motor synergies to achieve low-dimensional task-space tracking control of a human-like robotic arm actuated by redundant muscles, without prior knowledge of the robot. We propose to extract motor synergies from a subset of randomly generated reaching-like movement data. The essence is to first approximate the corresponding optimal control signals, using estimations of the robots forward dynamics, and to extract the motor synergies subsequently. In order to avoid modeling difficulties, a learning-based control approach is adopted such that control is accomplished via estimations of the robots inverse dynamics. We present a kernel-based regression formulation to estimate the forward and the inverse dynamics, and a sliding controller in order to cope with estimation error. Numerical evaluations show that the proposed method enables extraction of motor synergies for low-dimensional task-space control.

ROSlink: Interfacing legacy systems with ROS

This paper presents ROSlink, an open source project that aims at easing the integration of legacy systems with ROS (Robot Operating System). Its design principles provide a set of unique features that make it appealing for the interconnection of ROS with systems where ROS itself cannot be installed. First, ROSlink requires very limited changes to the legacy system. The project is self contained, bringing in no dependencies, which may be difficult to satisfy in a legacy system. Furthermore, with ROSlink any data type already in use in the legacy system can be employed for the communication of topics and service requests and responses. ROSlink allows run-time rerouting of the communication between the legacy system and ROS. Moreover, it empowers the legacy code with the ROS name remapping system, without enforcing any constraint on the command line parameters of legacy programs. Finally, by simply using a set of API that closely follow the ROS programming interface, ROSlink simplifies any successive porting of the code to a real ROS system. In this paper, the main design choices of ROSlink are discussed. A list of practical applications and tests where ROSlink was employed, as well as a short discussion on the projects future directions are then given.