Wireless Sensors: The Data You Want
无线传感器:满足你的一切数据需求
By Aaron Hand, Managing Editor
原文链接:http://www.controldesign.com/articles/2011/wireless-sensors-data-you-want.html
翻译:e-works王聪
校对:加州州立大学东湾分校 柯贤能(Ryan.Ke)
在日常生活与工作中,我们常常被庞大的数据淹没。上个月,思科的统一接入业务部副总裁Jeff Reed,在罗克韦尔自动化公司全球机械制造商论坛上就“zettaflood”现象警告到:现今我们面临的数据量不仅仅是百兆级别和千兆级别的了,现在数据的规模已经达到了PB级和EB级。毫无疑问数据收集模式的微小改变便可能造成机械运行模式发生变化。
“人们现在已经被数据淹没了。”霍尼韦尔传感与控制无线解决方案总监Todd Hanson很同意上述说法。“现在我们经常听到客户这样说:‘我虽然已经得到足够的数据了,但我仍然想要更多的数据。’如洪水般的数据并没有淹没他们渴望得到更多数据的欲望,他们需要更多的机器稳定运行时间、更多的生产力,因此他们需要传感器的数据。”
现在许多行业都非常抱有指望的是使更具有预测性的生产方式成为可能,而不是仅仅的被动的反应式接受,而这时就必须依仗工厂中的传感器接收而来的数据。处理这种对立和分歧的关键涉及的不仅仅是能否有更好的途径采集所有的数据,还于在提供客户所需要的数据时不至于使他们信息过量。“为了更好得对设备进行健康监控或是预测性的维护,他们需要大量的数据片段,” Hanson说,“他们需要这些可操作的数据。”
无线传感器的出现完美的解决了可控数据的需要,因为正如制造商追求的那样,无线传感器可以轻松的对现有的基础设施进行改造。“一个工厂可能已经拥有有线传感器,但有线的传感器可能并没有位于机械的轴承、压缩机泵或他们想要进行预测维护的其他任何地方。” Hanson说道,“在此类地方如果使用有线传感器将导致高昂的成本。”
在某些地方使用有线传感器并不实际,然而制造厂商对于数据点的检测需求却与日俱增。在美国威斯康星州的Waukesha Machine& Tool公司在CNC的工具加紧装配过程中,若用有线传感器来持续跟踪托盘之间缓慢的相互作用要求是很难达到。 “我们的客户需要一个无线压力监测系统,因为他们需要在任何给定的的时间内一台机器上运行多个托盘。”Waukesha Machine& Tool的设计师Kyle Spuhler说到。
运用有线传感器系统不能监控运动的机器应用状态。提供CNC数控系统等其他应用程序无线数控解决方案的电化学解决方案应用程序工程师Jim Stawitzky说道:“运用有线系统追踪运动的应用是会失败的。”
然而无线传感器可以远程遥控CNC数控中心。“如果无线传感器检测到一个制造过程中的压力损失,他将会发送给机器一个即停指令,替换掉现在的机器夹具机而改用另一个夹具,”Spuhler解释到。“它同样设立用来检测夹具在制造前的压力损失。如果出现此种情形,它会跳过此制造部分,转移到下一个。”
“无线传感器还能涉及另外一个问题---电池生命周期。”Stawitzky谈到,通常它都不会成为一个瓶颈,如果在使用时合理的维护以及循环的更换电池。
而事实上,有限的电池供电意味着无线传感器天生适合休息期期间的数据收集。“如果他们只想要一个预想的时间间隔的数据资料或数据包片段。”汉森说。
“其实客户只需要过程中或者可能发生变化的数据”Stawitzky补充道。“在一个拥有127个托盘的系统中,如果你只关注实时变化的数据的话,你只用观察其中小于1%的部分。”
对于机械制造商,无线传感器可以满足客户需要的这种灵活性。“对于一个设备供应商面临的困难是--我需要将传感器放在哪里才能监测到客户需要的数据?”汉森说,“在获得客户的同意后,他们会从布线或从机器的I / O接口处着手。如果不经协商直接布线交付给客户的话,难免会有错误的地方。”
无线传感器运用起来快捷方便。“他们轻易的从设备获得数据,配合控制器作为一个整体对设备进行控制;如果他们想,无线传感器甚至可以作为一个独立的元件,”汉森解释到。
In our lives and in our jobs, we are inundated with data. At Rockwell Automation's Global Machine Builder Forum last month, Jeff Reed, vice president of Cisco's Unified Access Business Unit, warned of the "zettaflood" of data—not gigabytes, terabytes, petabytes or even exabytes, but zettabytes. It's no wonder, then, that the mere mention of a system's data-gathering capabilities can turn machine operators running the other way.
"People are inundated with data now," agrees Todd Hanson, director of wireless solutions for Honeywell Sensing and Control (www.honeywell.com/sensing). "So we get customers that say, ‘Listen, I get enough data. The last thing I want is more data.' The flip side of that is that they want more data. They need higher productivity; they need more uptime. To get that, they need sensored data."
Many industries are looking to enable more predictive rather than reactive manufacturing, which requires the necessary data from a variety of sensors throughout a plant. Dealing with this dichotomy involves not only better ways to assimilate all the data, but also ways to give users the data they need without overloading them. "To get more equipment health monitoring or predictive maintenance, they want snippets of data," Hanson says. "They want actionable data."
Wireless sensors are a natural fit for actionable data because they can be easily retrofitted into an existing wired infrastructure to get just the data manufacturers are looking for. "A plant might already have wired sensors, but it may not have sensors that are located at the bearing or at the compressor pump or wherever they want to do predictive maintenance," Hanson says. "It can be cost-prohibitive to run wires to get that data."
In some environments, wired sensors just aren't practical, and yet there's a growing need for data points there as well. Waukesha Machine & Tool (www.wmtcinc.com) in Waukesha, Wis., makes workholding assemblies for CNC applications, where constantly moving pallets jogging against each other is too great a demand for wired systems. "Our customers require a wireless pressure monitoring system because of the multiple pallets they run in one machine at any given time," says Kyle Spuhler, designer for Waukesha Machine & Tool.
With wired systems, they just couldn't monitor this application, notes Jim Stawitzky, manager of application engineering for Electrochem Solutions (www.electrochemsolutions.com), which provides the wireless CNC solution for Waukesha, as well as other applications. "If something's in motion, wire is going to fail," he says.
But wireless sensors have even enabled lights-out manufacturing for CNC machining. "If the sensor detects a pressure loss during manufacturing, it will put the machine into an emergency stop and shuttle that fixture out of the machine and bring another fixture/part in for manufacturing," Spuhler explains. "It is also set up to determine if a fixture has a pressure loss before manufacturing, and if so, it will skip that part for manufacturing and move to the next."
One area where wireless sensors still can make strides, Stawitzky says, is in battery life. However, it typically has not been a hurdle, with battery replacement cycled in with scheduled maintenance.
And the fact that they have limited battery power available means that wireless sensors are an inherently good fit for gathering samples of data between resting periods. "If they're only wanting snippets of information or packets of data at a preconceived interval, this is what it does," Hanson says.
"Really, customers only need data when things are changing or in process," Stawitzky adds. "On a 127-pallet system, you can get less than 1% of the data if you look at just the pallet that's being machined."
For a machine builder, wireless sensors can enable the kind of flexibility that many customers need today. "For an equipment supplier, the challenge is where do I put that sensor to pull the data the customer may want," Hanson says. "They'll do their wiring and I/O layout on the machine, get approval from the customer, but once they deliver and go to commission, something's always wrong."
A wireless sensor can be dropped in quickly and easily. "They simply take the data from that device, tie it in to the overall controller if they want or even as a standalone," Hanson explains.
发表于:
2011-12-27 16:29 阅读(1207)
评论(3) 收藏 好文推荐
本博客所有内容,若无特殊声明,皆为博主原创作品,未经博主授权,任何人不得复制、转载、摘编等任何方式进行使用和传播。
作者该类其他博文: