Abstract
There is currently an increasing demand for robots able to acquire the sequential organization of tasks from social learning interactions with ordinary people. Interactive learning-by-demonstration and communication is a promising research topic in current robotics research. However, the efficient acquisition of generalized task representations that allow the robot to adapt to different users and contexts is a major challenge. In this paper, we present a dynamic neural field (DNF) model that is inspired by the hypothesis that the nervous system uses the off-line re-activation of initial memory traces to incrementally incorporate new information into structured knowledge. To achieve this, the model combines fast activation-based learning to robustly represent sequential information from single task demonstrations with slower, weight-based learning during internal simulations to establish longer-term associations between neural populations representing individual subtasks. The efficiency of the learning process is tested in an assembly paradigm in which the humanoid robot ARoS learns to construct a toy vehicle from its parts. User demonstrations with different serial orders together with the correction of initial prediction errors allow the robot to acquire generalized task knowledge about possible serial orders and the longer term dependencies between subgoals in very few social learning interactions. This success is shown in a joint action scenario in which ARoS uses the newly acquired assembly plan to construct the toy together with a human partner.

Abstract
The ability to employ autonomous vehicles to find and track the boundary between two different water masses can increase the efficiency in waterborne data collection, by concentrating measurements in the most relevant regions and capturing detailed spacial and temporal variations. In this paper we provide a guidance mechanism to enable an autonomous vehicle to find and track the steepest gradient of a scalar field in the horizontal plane. The main innovation in our approach is the mechanism to adapt the orientation of the crossings to the local curvature of the boundary, so that the vehicle can keep tracking the gradient regardless of its horizontal orientation. As an example, we show how the algorithms can be used to find and track the boundary of a dredged navigation channel, using only altimeter measurements. © 2014 IEEE.

Abstract

Abstract

Abstract
Wireless sensor networks are an emerging technology that can provide valuable information for a large series of monitoring and control applications. Comprising many monitoring scenarios with different particularities, as industrial management, weather forecasting, home automation, traffic management and rescue operations, just to cite a few, wireless sensor networks bring many possibilities for innovative applications that cannot be addressed by conventional wireless network technologies. When sensors are equipped with cameras or microphones, multimedia data can be retrieved from the monitored field, enriching the perception of the target area. However, the constrained nature of wireless sensor networks imposes many challenges to multimedia transmission, fostering development of optimized protocols. In this chapter, we present the state of the art of multimedia transmission in wireless sensor networks, covering topics as routing, error control, congestion avoidance, real-time delivery, compression and QoS, potentially supporting in the development of wireless multimedia sensor networks.

Abstract