December 3, 2016 | Written by: Dan Bigos
Share this post:
A recent Wall Street Journal article, ‘How U.S. Manufacturing Is About to Get Smarter,’ contrasts the promise of the IoT with the reality faced by many manufacturing organizations. The ability to realize IoT benefits varies by industry and age of equipment. With recently acquired, already-instrumented equipment and hi-tech manufacturing, you’re likely already realizing the benefits. For heavy industry, however, with older equipment lifetimes measured in decades, IoT benefits may be somewhere on the distant horizon. If you’re in the latter group, here are a few questions to consider to help plan your transition to a manufacturing environment that incorporates IoT. When possible we’ve provided links to resources for additional perspective.
What are your most critical assets?
Not all equipment has equal criticality, and it’s highly likely that the process of instrumenting numerous critical equipment will progress systematically. As a first pass, identify those assets that pose the greatest risk to productivity in the event of unplanned downtime. But how do you determine the proper criteria for assessing criticality? ‘How to effectively manage assets by criticality’ in Reliable Plant offers a straightforward method to rank assets by attributes to arrive at a ranked list of critical assets.
What are the best options for retrofitting older equipment with sensors?
The cost of sensors continues to decline. The variety of special purpose sensors continues to increase. Many sensors use batteries lasting months, even years, can be easily installed and use wireless communications to obviate the need for wiring or extensive equipment modification. The Wall Street Journal article ‘Bringing Smart Technology to Old Factories Can Be Industrial-Size Challenge‘ examines some of the challenges of retrofitting older equipment.
What are the best practices for acquiring data generated by newly-instrumented equipment?
Ideally, you want to minimize the physical connectivity in favor of wireless connectivity between equipment and the network. IBM partners Intel, Adlink and Cisco offer a wide range of solutions using edge technologies and gateways to securely and reliably make these connections, and in many instances bring analytic. ‘Smart Factories – the strategic role of wireless sensor networks‘ provides some practical and economic benefits of the wireless strategy.
Which operational data provides the greatest value and how much of it will I really need?
Of the myriad data you could potentially obtain from newly-instrumented equipment only a subset of the data, confirmed via careful analysis of data sources that correlate to predict equipment behavior, will likely be used. We’ve worked with clients who identified hundreds of potential data sources generated by very sophisticated equipment, yet when analyzed, determined that approximately 3 dozen of those sources were actually relevant in describing, monitoring, and predicting equipment behavior. Machine learning algorithms such as gradient boosting machines have simplified the process of sifting through volumes of operational data to correlate data that accurately monitors equipment behavior and prescribes appropriate remedies in the event of impending equipment failure. IBM’s Prescriptive Maintenance on Cloud offering, part of the IoT for Manufacturing portfolio, employs these algorithms to assess equipment reliability and prescribe optimum maintenance strategies.
Older equipment need not be a deterrent
Clearly there are many other aspects and situations unique to your organization to consider when planning to incorporate IoT technologies into manufacturing environments that employ older equipment. However, equipment age should not be a deterrent. As evidenced by the aforementioned articles, manufacturers across many industries are successfully adopting IoT and realizing benefit.