Continuous monitoring of production of yarn is an important process requirement for good productivity in a ring frame machine in textile industry. So, collecting the data on yarn breakages using an appropriate Wireless Sensor Network (WSN) has been taken as a research problem in this work. The paper describes a method in which a mobile agent communicates wirelessly with the sensor network to collect the yarn breakage data which would serve the mobile end-user. With a view to reducing the energy consumption of the sensors and the end-to-end delay, a new approach, viz., Mobile Agent-Based Event-Driven Wireless Sensor Network (MAEDWSN), has been proposed. In this scheme, a new residual energy-based cluster creation algorithm for dynamic selection of cluster head has been proposed to save the energy constrained sensor nodes. An attempt has been made using mobile agent middleware called Java Agent DEvelopment framework (JADE) which facilitates the agent development. The collected yarn breakage data by the mobile agent has also been made available to the end-user with the help of ‘Android’ a virtual mobile device. The simulated results have been compared with the well-known client-server approach.

Over the past three centuries, spinning technology has been getting continuously improved through thousands of minor innovations, and occasional major advances that have collectively increased the quality and lowered the cost of producing yarn. In spinning mills, the most important parameter to be monitored while yarn is being formed is the yarn breakage. Hence, detecting the yarn breakage in a spinning machine in a textile industry is one of the important tasks as it directly affects the production efficiency. In current sensing technique, a wired network is used in the spinning machine. Wireless communication, if used, can however remedy the cabling issues with the traditional monitoring system usage and thereby significantly reduce the maintenance costs (Stuart et al., 2009).

Wireless Sensor Network (WSN) finds ameliorating popularity in industrial sensing applications due to its comparatively low cost and simplicity for improvement into existing infrastructure (Baker et al., 2009). Progressive improvements in the hardware technology have resulted in manufacturing miniature sensor nodes on single chips with embedded memory, processor, transceiver and a companion battery. As it is difficult to replace or charge the exhausted battery, this aspect gives way to the primary objective of increasing the lifetime of the battery, leaving the optimization of other performance metrics as the second objective. Since the communication aspect involved in the sensor node is more energy-consuming than their involvement in computational aspects (Linnyeret al., 2004; and Mohammad and Imad, 2005), it becomes mandatory to keep the communication duration as minimum as possible, so it would be better to collect the desired data from the sensor networks directly rather than communicating through the nearby sink. In this context, yarn breakage data collection using an appropriate WSN, in addition to supporting real-time dynamic configuration of collection of tasks, seems to be a challenging research. The key performance metrics in WSN are both network lifetime and an average time required to report an event reliably (Fatma et al., 2008). Taking the two aspects, viz., yarn break detection and the subsequent data collection, it is also felt that a mobile agent-based framework if used for collecting data on yarn breakages through the use of an appropriate sensor network would also be beneficial. With this in view, in the proposed approach, the sensed information has been collected using mobile agents, and it has been found that the energy spent on sensor nodes getting minimized.

Information Technology Journal, WSN, Mobile agent, JADE, Android.