Dit artikel is met toestemming overgenomen uit Microniek, 2020, nr 5 Robotics research groups around the world are using Robot Operating System (ROS)to develop their prototypes quickly. While the first version of ROS was aimed primarilyat the R&D community, its successor, ROS 2, has been redesigned completely to beindustrial grade and applicable in research, prototyping, deployment and production.This allows ROS 2 prototypes to evolve into products suitable for real-worldapplications. To explore the state of the art, Saxion University of Applied Sciencesand nine companies are developing an industrial mobile robot. This article describesexperiences from the development process and presents an outlook on the potentialof ROS 2 for industry.
MULTIFILE
Abstract Long-haul travel does not constitute an obstacle for tourists to travel and is fast gaining the attention of tourists in new and unique experiences. This study was conducted to identify the long-haul travel motivation by international tourists to Penang. A total of 400 respondents participated in this survey, conducted around the tourist attractions in Penang, using cluster random sampling. However, only 370 questionnaires were only used for this research. Data were analysed using SPSS software 22 version. The findings, ‘knowledge and novelty seeking’ were the main push factors that drove long-haul travel by international tourists to Penang. Meanwhile, the main pull factor that attracts long- haul travel by international tourists to Penang was its ‘culture and history’. Additionally, there were partly direct and significant relationships between socio-demographic, trip characteristics and travel motivation (push factors and pull factors). Overall, this study identified the long-haul travel motivations by international tourists to Penang based on socio-demographic, trip characteristics and travel motivation and has indirectly helped in understanding the long-haul travel market particularly for Penang and Southeast Asia. This research also suggested for an effective marketing and promotion strategy in pro- viding useful information that is the key to attract international tourists to travel long distances. Keywords:
Autonomous robots require high degrees of cognitive and motoric intelligence to come into our everyday life. In non-structured environments and in the presence of uncertainties, such degrees of intelligence are not easy to obtain. Reinforcement learning algorithms have proven to be capable of solving complicated robotics tasks in an end-to-end fashion without any need for hand-crafted features or policies. Especially in the context of robotics, in which the cost of real-world data is usually extremely high, reinforcement learning solutions achieving high sample efficiency are needed. In this paper, we propose a framework combining the learning of a low-dimensional state representation, from high-dimensional observations coming from the robot 's raw sensory readings, with the learning of the optimal policy, given the learned state representation. We evaluate our framework in the context of mobile robot navigation in the case of continuous state and action spaces. Moreover, we study the problem of transferring what learned in the simulated virtual environment to the real robot without further retraining using real-world data in the presence of visual and depth distractors, such as lighting changes and moving obstacles.
MULTIFILE
Even though considerable amounts of valuable wood are collected at waste collection sites, most of it remains unused and is burned: it is too labor-intensive to sort, process and upcycle useable parts. Valuable wood thus becomes worthless waste, against circular economy principles. In MoBot-Wood, waste collection organizations HVC and the municipality of Amsterdam, together with Rolan Robotics, Metabolic and AUAS investigate how waste wood can be sorted and processed at waste collection sites, using an easy-to-deploy robotic solution. In various preceding and on-going projects, AUAS and partners are exploring circular wood intake, sorting and processing using industrial robots, including processes like machine vision, 3D scanning, sawing, and milling. These projects show that harvesting waste wood is a challenging matter. Generally, the wood is only partially useable due to the presence of metal, excessive paint, deterioration by fungi and water, or other contamination and damages. To harvest useable wood thus requires intensive sorting and processing. The solution of transporting all the waste wood from collection sites to a central processing station might be too expensive and have a negative environmental impact. Considering that much of collected wood will need to be discarded, often no wood is harvested at all, due to the costs for collection and shipping. Speaking with several partners in related projects, the idea emerged to develop a mobile robotic station, which can be (temporarily) deployed at waste collection sites, to intake, sort and process wood for upcycling. In MoBot-Wood, research entails the design of such station, its deployment conditions, and a general assessment of its potential impact. The project investigates robotic sorting and processing on location as a new approach to increase the amount of valuable, useable wood harvested at waste collection sites, by avoiding material transport and reducing the volume of remaining waste.
