Industrial breakdowns are costly, as are their close cousins: unplanned maintenance to prevent an imminent breakdown. There is a body of evidence that suggests that continuously monitoring industrial equipment to detect early signs of performance degradation or failure is a worthwhile process that will help mitigate lost productivity, expensive repairs, and process inefficiency.
While it’s a great theory, many companies can’t monitor effectively due to aging equipment and a lack of integration of the limited analytics they do have. To try and overcome these challenges, IBM and National Instruments (NI) are working on a test bed project with the goal of developing new predictive maintenance analytics modeling techniques and creating standards for a modern industrial networking environment. Both are members of the Industrial Internet Consortium.
The ultimate goal of the test bed, they say, is to produce a multi-vendor, cloud-based predictive maintenance solution unlike existing platforms. While it operates, the Condition Monitoring and Predictive Maintenance Testbed will engage in continuous online measurements, automated analysis, and balance of plant coverage. IBM will provide the cloud environment and analytics, and NI will supply the monitoring and data acquisition technology.
“To bring industrial equipment up to date by today’s standards, modifications need to be made so that these machines are able to fully harness the potential of the Internet of Things (loT) to provide condition monitoring solutions,” NI’s Stuart Gillen, condition monitoring platform vibration analyst, told Design News. “Combining sensor data from multiple pieces of equipment and/or multiple processes can provide deeper insight into the overall impact of faulty or sub-optimal equipment in advance of catastrophic and costly repairs.”
One of the biggest challenges of the testbed will be to demonstrate the frontiers of analytics as they apply to predictive maintenance. The more data analysis, the better a company can correlate machine data with operational data so maintenance can be optimally scheduled around production requirements. The new solution will be available to new equipment, but is also being geared toward older machinery for the purpose of retrofitting.
“Some of the systems being used in industry for condition monitoring and predictive maintenance today are not open and can’t adapt to all the new sensors being created in the loT,”said Gillen. “NI’s InsightCM software is open to implement new capabilities as needed, and IBM’s Bluemix platform is wide open—it can even handle the interplay of non-traditional data from sources like Twitter and weather forecasts to correlate with production or operations data.”
The sources will ultimately be pulled together by a host of networking technologies, including ZigBee, Bluetooth, Ethernet, and more. Since security is a prime concern in IoT installations, the project will use a multilayer approach, according to Greg Gorman, IBM’s director for Internet of Things.
“The first consideration is just the basic authorization of whether or not the device should be allowed to connect,” Gorman told Design News. “Then, encrypting data to keep it from prying eyes. Also important is ensuring the device isn’t being ‘faked’; that the device is who it says it is. Finally, end-to-end access control to ensure that people can only access the data from the device (or some of the data) that they are authorized to see,” he said.
IBM, for its part, eventually hopes to go beyond existing data analytics techniques to help the resulting solution gain traction over existing industrial monitoring platforms. While phase one of the test bed will use IBM’s existing analytics technology, PMQ and Maximo, later phases will include advanced technology from Watson, the company’s artificial intelligence system. Gorman says it’s about helping companies do more with less.
“Specialists on machines are expensive to develop, so making the most effective use of them is key,” he said. “Also, just the cost savings from doing predictive maintenance justifies the whole push in this direction — efficient use of all the resources of the organization — humans and machines.”