6533b85efe1ef96bd12c0660

RESEARCH PRODUCT

Pilot Prototype of Autonomous Pallets and Employing Little’s Law for Routing

subject

description

Application of autonomous control for shop-floor scheduling by considering real-time control of material flows is advantageous to those assembly lines with dynamic and uncertain circumstances. Among several potential processors with computing and communication capabilities—for representing autonomous material carriers—wireless sensor nodes seem as promising objects to be applied in practice. For realizing autonomy in making scheduling and routing-control decisions some methodologies need to be embedded in the nodes. Among several experimented methodologies, e.g., artificial intelligence, genetic algorithm, etc., in the context of a doctoral research, in this current special case of assembly scenario, the queuing theory and its simple equations seem quite suitable. For instance, employment of Little’s law for calculating and analysis of simple queuing structures is a favorable method for autonomous pallets in real shop-floors. Concerning the simplicity and inexpensive computing loads of such a rule, it suits the best to the low capacity wireless sensors in developing pilot prototypes of autonomous carriers. Little’s law can be used to estimate the current waiting times of alternative stations and try to find a non-decreasing order of operations to improve the performance record (e.g., makespan) of the entire assembly system. To develop a pilot prototype, some wireless sensors—representing pallets in practice—are connected to a simulated assembly scenario via the TCP/IP protocol to evaluate the feasibility of realizing autonomous pallets in the practice of shop-floor control. Nevertheless, wireless nodes are distributed objects, so the use of data sharing for transferring low data between each other and respectively low energy consumption is necessary.