How the concept of asset management is evolving with the Industrial Internet of Things to address assets ranging from robots to I/O blocks
The cutting edge thought process in manufacturing today is to use the automation network for more than just automation. That could mean using it for safety, energy management, embedded web servers or asset management. Asset management may be considered “old hat” in the manufacturing space, but under the influence of the Industrial Internet of Things (IIoT), new life is being breathed into the topic.
What’s new is the extent to which asset management is evolving. It’s not just about factory lines or cells or even individual machines anymore—it’s now coming down to singular devices.
There are two main drivers behind this development. The first is the installation of Ethernet into the farthest reaches of the network. This allows asset management information to be retrieved transparently. The second is the increased ease with which data can be mined, stored and analyzed. This means complete information on a device can be gathered, not just fragmented pieces.
The benefits of asset management for connected devices as a profit driver are well known, but to achieve this you must leverage the information stored in memory on the device. This information falls in two categories: identification data (I-data) and maintenance data (M-data). Together, this I&M data contains all the required information needed to label, locate and describe a networked device. The device can be anything: a robot, a motor, a valve or an I/O block. I-data records contain read-only information about the device like the serial number, hardware version and software version. M-data on the other hand, are read-write data written at the time of installation, such as a plant-wide unique identifier, location or installation date.
The best way to understand the usefulness of I&M data is via example. Consider a network with multiple instances of the same type of robot. These robots may have been installed at different times or upgraded along the way. Performing an I&M audit quickly informs a maintenance manager that some of the robots do not have the latest firmware version installed. This is particularly useful in large networks, where I-data makes it easy to distinguish similar devices from each other. In general, it’s best to think of I&M data as a traditional nameplate, but smarter and in electronic form.
M-data is not only written during device installation, it is also acquired during production. Critical components can trigger a flag when certain preset maintenance conditions are met. Typically this involves an extended traffic-light model to display device states. In addition to the standard failure state, pre-warning levels such as “maintenance required” and “maintenance demanded” are also employed. For example, if a plant manager knows the valve will degrade after 10,000 actuations, a “maintenance required” flag is set after 8,000, and a “maintenance demanded” flag is set after 9,000. This can help eliminate costly unplanned downtime.
Many of these concepts are not new, having been part of the Profinet specification for a decade. In the ISA-95 hierarchy, I&M data is passed from the field level to the MES level and to the ERP level. With many IIoT futurists predicting a transformation of that hierarchy into a flatter network architecture, the connection from field level systems to MES and ERP systems will become more transparent and direct. Therefore the focus of asset management needs to expand to include the network to which those devices are connected.
Because a factory’s automation network is used for more than just automation, it is an infrastructure and an asset unto itself. Just like I&M data is used to manage the assets connected to a network, there are tools available to manage the network as an asset. Due to the proliferation of Ethernet at the office level, its adoption at the field level allows manufacturers to leverage many of the tools developed for it along the way.
One such tool is the Simple Network Management Protocol (SNMP). It can be used to query Ethernet switches (standalone, or in devices) for many Ethernet-related parameters like bandwidth utilization, retries, link status, etc. Another tool is Link Layer Discovery Protocol (LLDP), which makes it easy to extract the topology of a network. Finally, Hyper Text Transfer Protocol (HTTP)—which is so universal that we almost forget it’s a protocol—delivers the ability to browse web pages in connected devices.
Article appears in the August issue of AutomationWorld magazine