Abstract

Abstract
The world is living the fourth industrial revolution, marked by the increasing intelligence and automation of manufacturing systems. Nevertheless, there are types of tasks that are too complex or too expensive to be fully automated, it would be more efficient if the machines were able to work with the human, not only by sharing the same workspace but also as useful collaborators. A possible solution to that problem is on human-robot interaction systems, understanding the applications where they can be helpful to implement and what are the challenges they face. This work proposes the development of an industrial prototype of a human-machine interaction system through Augmented Reality, in which the objective is to enable an industrial operator without any programming experience to program a robot. The system itself is divided into two different parts: the tracking system, which records the operator's hand movement, and the translator system, which writes the program to be sent to the robot that will execute the task. To demonstrate the concept, the user drew geometric figures, and the robot was able to replicate the operator's path recorded.

Abstract
Understanding the distribution of people's transportation mode is a crucial facet of today's urban mobility for proper transportation planning. The penetration of smartphones combined with their sensing capability is an enabler for crowdsourcing large mobility data such as commuters' GPS records. In this paper, we leverage the GPS traces of commuters to infer five different transportation modes frequently used in urban areas including foot, bike, bus, car and metro. We compare three different approaches commonly reported in the literature for transportation mode detection from the family of machine learning algorithms (random forest -RF) and deep learning architectures (convolutional neural network -CNN and ensemble of autoencoders -EAE). By splitting the dataset into train-test by the period of data collection, as well as the conventional 80-20 split, we evaluate the impact of several data pre-processing decisions on overall classifiers' performance. Our results show RF and CNN performing better upon evaluation on classification metrics such as the f1 score and the area under the Receiver Operating Characteristics (ROC) curve.

Abstract
This experimental study focuses on the comparison between two different sensors for vibration signals: a magnetoresistive sensor and an accelerometer as a calibrated reference. The vibrations are collected from a variable speed inductor motor setup, coupled to a ball bearing load with adjustable misalignments. To evaluate the performance of the magnetoresistive sensor against the accelerometer, several vibration measurements are performed in three different axes: axial, horizontal and vertical. Vibration velocity measurements from both sensors were collected and analyzed based on spectral decomposition of the signals. The high cross-correlation coefficient between spectrum vibration signatures in all experimental measurements shows good agreement between the proposed magnetoresistive sensor and the reference accelerometer performances. The results demonstrate the potential of this type of innovative and non-contact approach to vibration data collection and a prospective use of magnetoresistive sensors for predictive maintenance models for inductive motors in Industry 4.0 applications.

Abstract
The automation of internal logistics and inventory-related tasks is one of the main challenges of modern-day manufacturing corporations since it allows a more effective application of their human resources. Nowadays, Autonomous Mobile Robots (AMR) are state of the art technologies for such applications due to their great adaptability in dynamic environments, replacing more traditional solutions such as Automated Guided Vehicles (AGV), which are quite limited in terms of flexibility and require expensive facility updates for their installation. The application of Artificial Intelligence (AI) to increase AMRs capabilities has been contributing for the development of more sophisticated and efficient robots. Nevertheless, multi-robot coordination and cooperation for solving complex tasks is still a hot research line with increasing interest. This work proposes a Multi-Agent System for coordinating multiple TIAGo robots in tasks related to the manufacturing ecosystem such as the transportation and dispatching of raw materials, finished products and tools. Furthermore, the system is showcased in a realistic simulation using both Gazebo and Robot Operating System (ROS).

Abstract
The electric vehicle (EV), when aggregated by an agent (Aggregator), is a suitable candidate for participating in demand response in power system operation. As the interface between distribution network and EV users, as well as an independent party at the same time, an optimal scheduling algorithm is necessary with consideration of benefits of three parties, which in return will affect aggregators' sustainable development. The benefits of distribution system from demand response, aggregator and EV users are defined in this paper. EV users' benefit is described by their satisfaction on SOCs reached after a given period of time and overall costs/revenues for charging/discharging and policy award/penalty, while the benefit of distribution network for the integration of large amount EV loads through aggregator is evaluated by aggregator's load shifting capability through a pricebased demand response (DR) program under real time electricity price. The optimal scheduling of the aggregator is with an objective of maximizing its own benefit under constraints of EV users' minimum satisfaction and minimum load-shifting capability required by distribution network. The optimization scheduling is tested by a test system, and further analysis is given on the effect of aggregator's facility level and technology (Vehicle to Vehicle) and the operation mode of aggregator group on the benefits of three parties.