This study proposes a new model for guiding people in urban settings using multiple robots that work cooperatively. More speci cally, this investigation describes the circumstances in which people might stray from the formation when following di erent robots' instructions. To this end, we introduce a \Prediction and Anticipation Model" that predicts the position of the group using a Particle Filter, while determining the optimal robot behavior to help people stay in the group in areas where they may become distracted. As a result, this article presents a novel approach to
locally optimizing the work performed by robots and people using the minimum robots' work criterion and determining humanfriendly types of movements. The guidance missions were carried out in urban areas that included multiple conflict areas and obstacles. This study also provides an analysis of robots' behavioral reactions to people by simulating di erent situations in the locations that were used for the investigation.
The method was tested through simulations that took into account the di culties and technological constraints derived from real-life situations.
Despites these problematic issues, we were able to demonstrate the robots' e ect on people in real-life situations in terms of pushing and dragging forces.
Despite the significant advances in path planning methods, highly constrained problems are still challeng-
ing. In some situations, the presence of constraints defines a configuration space that is a non-parametrizable
manifold embedded in a high dimensional ambient space. In these cases, the use of sampling-based path
planners is cumbersome since samples in the ambient space have low probability to lay on the configuration
space manifold. In this paper, we present a new path planning algorithm specially tailored for highly constrained
systems. The proposed planner builds on recently developed tools for higher-dimensional continuation, which
provide numerical procedures to describe an implicitly defined manifold using a set of local charts. We propose
to extend these methods focussing the generation of charts on the path between the two configurations to connect
and randomizing the process to find alternative paths in the presence of obstacles. The advantage of this planner
comes from the fact that it directly operates into the configuration space and not into the higher-dimensional
ambient space, as most of the existing methods do.
Aksoy , E.; Abramov, A.; Dörr, J.; Ning, K.; Dellen, B.; Wörgötter , F. International journal of robotics research Vol. 30, num. 10, p. 1229-1249 DOI: 10.1177/0278364911410459 Data de publicació: 2011-10-28 Article en revista
Recognizing manipulations performed by a human and the transfer and execution of this by a robot is a difficult problem. We address this in the current study by introducing a novel representation of the relations between objects at decisive time points during a manipulation. Thereby, we encode the essential changes in a visual scenery in a condensed way such that a robot can recognize and learn a manipulation without prior object knowledge. To achieve this we continuously track image segments in the video and construct a dynamic graph sequence. Topological transitions of those graphs occur whenever a spatial relation between some segments has changed in a discontinuous way and these moments are stored in a transition matrix called the semantic event chain (SEC). We demonstrate that these time points are highly descriptive for distinguishing between different manipulations. Employing simple sub-string search algorithms, SECs can be compared and type-similar manipulations can be recognized with high confidence. As the approach is generic, statistical learning can be used to find the archetypal SEC of a given manipulation class. The performance of the algorithm is demonstrated on a set of real videos showing hands manipulating various objects and performing different actions. In experiments with a robotic arm, we show that the SEC can be learned by observing human manipulations, transferred to a new scenario, and then reproduced by the machine.
This paper presents a new method to solve the
configuration problem on robotic hands:deter-
mine how a hand should be configured so as to
grasp a given object in a specific way, characterized by a number of hand-object contacts to be satisfied. In contrast to previous algorithms given for the same purpose, the one presented here allows
specifing such contacts between free-form regions on the hand and object surfaces, and always returns a solution whenever one exists. The method is based on formulating the problem as a system of polynomial equations of special form, and then
exploiting this form to isolate the solutions, using a numerical technique based on linear relaxations. The approach is general, in the sense that it can be applied to any grasping mechanism involving lower-pair joints, and it can accommodate as many hand-object contacts as required. Experi-
ments are included that illustrate the performance of the method in the particular case of the Schunk Anthropomorphic hand.
Rodriguez, A.; Basañez, L.; Colgate, E.; Faulring, E. International journal of robotics research Vol. 29, num. 4, p. 336-352 DOI: 10.1177/0278364909104841 Data de publicació: 2010-04 Article en revista