Dynamic Scheduling Nears

The topic of dynamic scheduling caught my attention when it came up in a Webinar recently presented by Mazak. Titled “MTConnect—Downstream Applications to Increase Productivity,” the Webinar was moderated by Neil Desrosiers, an applications engineer/developer at Mazak and the company’s lead MTConnect specialist. He did a fine job of introducing MTConnect and clearly outlining how it functions as a plug-and-play networking platform for manufacturing applications.  

Neil opened his presentation by characterizing MTConnect as an open-source, royalty-free manufacturing protocol that easily connects devices and systems from different suppliers to capture and share information in a common format such as XML. After discussing the basic nuts and bolts of how this protocol is implemented on machine tools and connected to valuable applications such as production monitoring, Neil pointed to some of the enticing possibilities now lying before us. Dynamic scheduling is one of these. He explained that dynamic scheduling is an outgrowth of applications using MTConnect to facilitate communications between manufacturing devices such as a turning center and the attached bar feeder. Although the I/O systems of these devices can be interfaced to automate their joint operation, MTConnect enables the connection to exploit software programs that intelligently adjust a programmed sequence of turned parts.  

For example, it is possible to “nest” a series of parts so that the turning center produces a kit of related components for an assembly. By monitoring bar length, the system can determine when remaining stock is inadequate to produce one more nested kit. At that point, the system adjusts the sequence of part programs to machine those for which enough stock is still available. This optimizes production and wastes less stock. 

Handling this communication via a readily networked data format enables higher levels of centralized production monitoring and instant adjustment. Any number of turning centers and bar feeders could report and coordinate production among them. The next obvious step is to expose all machines in a shop to data in the scheduling system so that upcoming jobs can be shuffled about appropriately with little delay. 

This concept requires that the machines “talk the same language” to share critical information about on-board cutting tools, part identifiers, workpiece material characteristics and so on to match capability and availability to a “hit list” downloaded from the scheduling system. MTConnect is well on its way to establishing all of the naming conventions and data tags enabling this to happen. Neil believes that a prototype system will be technically feasible by the end of 2013. 

Of course, a growing number of applications for machine monitoring, energy management and tool control are already fully commercialized. The best place to find out about current and envisioned applications for MTConnect will be at the MTConnect: Connecting Manufacturing Conference, April 10-11, 2013 in Cincinnati.