In this work, the design of a multi agent system (MAS) for inventory control at the main Spanish e-commerce retailer ‘Group Lo Monaco’ is presented, which will contribute to improve the picking process at the warehouse. The proposed system uses Radio-Frequency Identification (RFID) technology for product identification. A MAS is used to place orders to suppliers and the agent methodology is used for generating mixed orders to clients. This paper introduces the state of the art for the proposed technologies, then proceeds to explain operations taking place at the company and whether the proposed technologies can be used in this case or not. Finally, a specific solution is proposed and conclusions are extracted.
1 April 2012 in Engineering Applications of Artificial Intelligence
Energy saving has become a major concern for the developed society of our days. This paper presents a Wireless Sensor and Actuator Network (WSAN) designed to provide support to an automatic intelligent system, based on the Internet of Things (IoT), which enables a responsible consumption of energy. The proposed overall system performs an efficient energetic management of devices, machines and processes, optimizing their operation to achieve a reduction in their overall energy usage at any given time. For this purpose, relevant data is collected from intelligent sensors, which are in-stalled at the required locations, as well as from the energy market through the Internet. This information is analysed to provide knowledge about energy utilization, and to improve efficiency. The system takes autonomous decisions automatically, based on the available information and the specific requirements in each case. The proposed system has been implanted and tested in a food factory. Results show a great optimization of energy efficiency and a substantial improvement on energy and costs savings.
30 November 2006 in Computers & Industrial Engineering
Recent research on highly distributed control methods has produced a series of new philosophies based on negotiation, which bring together process engineering with computer science. Among these control philosophies, the ones based on multiagent systems (MASs) have proved to be especially suitable to address distributed decision making in highly volatile environments. Furthermore, this agent-based approach allows using new manufacturing concepts in scheduling problems because of its scalability, ability to negotiate, and distributed capabilities. However, MAS models have the drawback of excessive dependence on up-to-date field information. In this paper, a theoretical and experimental MAS is presented to test new strategies for managing handling operations supported by radio-frequency identification information feedback. Based on this new visibility framework, these strategies applied to scheduling control have been structured on a new distributed organization model to enforce the idea of the division between physical elements and information and communication technologies. This division into two platforms simplifies the design and development of the MAS, providing a higher level of abstraction and preserving the independence between platforms. The proposed MAS, called MAS-DUO, has been validated using a test bench simulator and at the Ciudad Real Central Airport in Spain.