1660343828585 Eitan

Top 10 IIoT asset-maintenance trends

Oct. 16, 2017
A look ahead at insights from leaders (and laggards) in manufacturing. 

2017 was not as eventful a year as many had expected for IIoT asset-maintenance. Although almost every industrial manufacturer has included it as an operational or strategic priority,

Presenso's Eitan Vesely

some are holding back on major investments until there is more clarity in the market. At the same time, others are rushing forward.

The technology outlook is rapidly developing. Market entrants battle entrenched vendors that are upgrading their solution-offerings to keep pace with the new entrants’ innovation.  

Let's look to the future and share insights from our commercial engagement with leaders (and laggards) in manufacturing. 

Trend No. 1—Pause in IIoT infrastructure investment

In early 2017, there was no sign of letup in IIoT R&D investments on the part of industry behemoths such as GE and Siemens. However, GE has announced a strategic a shift to cost-cutting and a slowdown in investment and customer acquisition. What does this indicate for the overall predictive-analytics segment? GE has not achieved the growth within industrial plants that analysts had expected and this indicates a hesitancy on the part of the industrial sector to commit to a singular IIoT platform. 

Trend No. 2—Momentum for unsupervised machine learning (ML)

With Unsupervised Machine Learning, advanced algorithms analyze machine sensor data without the need to “train” the data labels. Whereas supervised ML requires the learning algorithm to be trained on the physical machine blueprints and mechanical processes, unsupervised ML is agnostic to vendor, asset age or sensor type. With advances in unsupervised ML, industrial plants now have the opportunity for this low-touch analytics methodology. 

Trend No. 3—Automotive industry will become the clear leader in IIoT asset-maintenance

More than any other sector, the automotive industry recognizes the potential value from IIoT predictive maintenance from both strategic and operational perspectives. The significant investments in R&D touch all aspects of the manufacturing process, including a serious commitment to reduce the industry’s Achilles heel—unscheduled downtime. The automotive companies that demonstrated a willingness to be early adopters of nascent technologies are expected to be the first to benefit financially from advances in AI and advanced predictive analytics.

Trend No. 4—IIoT predictive maintenance will be viewed as a source for top-line growth

The typical cost-justifications for traditional predictive maintenance (PdM) are based on increased operational efficiencies and savings. With Industry 4.0, executives are starting to consider the impact on top-line revenue from their big-data investments. With the shift from Industry 3.0 to Industry 4.0, metrics such as improved uptime and higher-production yield rates are replacing downtime as the driving force for investments in this technology category.

Do you understand prescriptive maintenance? Dig into our new Technology Report here. 

Trend No. 5—Holistic view of predictive maintenance within IIoT asset maintenance

In parallel to the growing acceptance of IIoT predictive maintenance, there are advances in other asset-maintenance processes. Within the context of Industry 4.0, industrial plants are adopting practices to automate and improve repair scheduling, inventory management and inspection. We are seeing a move away from siloed, project-driven approaches to IIoT asset maintenance, and the embrace of a holistic view of these complimentary solutions.

Trend No. 6—Big data centers of excellence will lose favor

In the past, executives have raised concerns about the inability of most industrial plants to recruit big data engineers and scientists. The underlying assumption is that in the race to Industry 4.0 adoption, big data centers of excellence are required. With the rapid pace of innovation, industrial plants are recognizing that 1) It is unrealistic to build these competencies internally and 2) Third-party vendor solutions will need to build professional services or automated functionality into their solution offerings.

Trend No. 7—Shift to multi-asset predictive-maintenance solution offers

In the past, predictive-maintenance solutions based on SCADA were sensor- or machine-specific. With the application of big-data analytics, industrial plants are now looking for single multi-asset predictive maintenance solutions that are not limited to a specific vendor or asset class.

Trend No. 8—OEM’s moving to service model

Traditional OEM’s have recognized the economic potential from the recurring-revenue model typically associated with software as a service (SaaS) licensing agreements. Many OEM’s are actively exploring this new model and we expect in 2018 that hardware vendors will announce that they will bundle their products with predictive maintenance and other service offerings

Trend No. 9—New players will emerge in the vendor ecosystem

The IIoT asset-maintenance category is attracting significant investment capital. Traditional vendors are investing in or acquiring startups, or building internal capabilities. In any revolution, new elites gain power at the expense of existing institutions. The fourth industrial revolution is no different. Even if new Googles and Amazons do not emerge in the industrial sector, numerous startups are driving new areas of IIoT innovation. 

Trend No. 10—Industry analysts will lose credibility

No need to name names. A new cottage industry of analysts is now riding the IIoT wave. As a consummate reader of research reports and whitepapers, I am astounded at the atmospheric growth predictions for IIoT solutions. In the real world, industrial plants are addressing the strategic implications of disruptive change while managing their day-to-day operations. As we get closer to 2020 (the end-year for 2015 predictions), we will find that past predictions of multi-billion-dollar IT categories will be shown to have been exaggerated. 

Eitan Vesely is CEO of Presenso