Talk to any machine tool maker about their latest offerings and the synonyms “intelligent” and “smart” come up a lot. Everyone is pushing the idea that their products are not just faster and stronger than the next guy’s but smarter as well; that they can essentially think for themselves. But has anyone truly achieved an intelligent machine tool? John Kohls doesn’t think so. “We don’t have an intelligent machine yet, but we have pieces of it,” he says, referring to the various smart functions touted by machine makers. And he wants to put those pieces together to make a new generation of machines that would live up to the hype and actually be able to think for themselves. To that end, Kohls, executive vice president of TechSolve Inc. (www.techsolve.org; Cincinnati, OH), a manufacturing consulting firm specializing in advanced machining processes and lean techniques, has become one of the driving forces behind the Smart Machine Platform Initiative (SMPI). SMPI was started by a consortium of organizations including the Association for Manufacturing Technology (www.mfgtech.org) and the Society of Manufacturing Engineers (www.sme.org.)
While there are machines that can sense a broken tool and discontinue operation, that’s not “intelligent,” according to Kohls: “That’s just a planned reaction to a sensor input as opposed to thinking like a person would.” He says true intelligence would compensate for several factors simultaneously and learn from the situation. What SMPI members envision is a machine tool that would:
How feasible is all of this? Up to now SMPI has basically laid the groundwork by determining what it wants, which is arguably the easy part. The next step is to construct a test bed using existing machines and as many intelligent bits of hardware as the group can afford, which is scheduled to be in place by the end of 2005. After that, the plan is to start validating and commercializing newly developed hardware and software that conforms to SMPI’s common standard. Kohls lays out the logic that drives this piecemeal approach: “Different machine tool builders will focus on different pieces of the smart machine for their specific applications. But the advantage is that the standard for interoperability will have been set.” The challenge to SMPI is to come up with technology so compelling that machine tool companies will suspend their traditional reluctance to join open standards and participate. If the initiative can achieve that then the next generation of machine tools could make parts markedly better, faster and cheaper. That’s smart.
Many engineering disciplines have come to rely heavily on computer simulations to save development time and money, but the practice has yet to catch on big in machining. In fact, Nishant Saini, sales manager, Third Wave Systems (Minneapolis, MN; thirdwavesys.com), estimates that fewer than 25% of machining operations are simulated before cutting a new part. Why such a low percentage? In part, it’s a combination of lack of familiarity with sophisticated software on the shop floor and a dearth of solutions aimed specifically at machining.
That’s where Third Wave Systems comes in. The company makes two machining simulation modules. One is an FEM (finite element method) solution called AdvantEdge aimed at part and process designers that employs capabilities like tool wear prediction, automated 3D residual stress analysis and 3D tool thermal boundary conditions to help engineers visualize and adjust processes virtually, decreasing the need for physical prototyping. According to Saini, the difference between AdvantEdge and many other FEM packages is that the code has been tailored for machining conditions, such as. the structural analysis of extremely high deformation associated with metal removal. The other package, AdvantEdge Production Module, is more for use on the plant floor and focuses on manufacturing concerns like cycle time calculation and the force and power predictions needed to improve machine utilization.
While it may seem that machining simulation is for larger companies only, Saini estimates that smaller automotive operations could cut tooling costs by 15 to 20% by switching from physical trials to simulations. Add that to reduced scrap costs and shorter process development times and the potential savings could be big enough for even software-adverse manufacturers to take a look at going virtual.