Whether you call it smart manufacturing or Industry 4.0, digital disruption is real. It is going to impact all manufacturers; asset-intensive businesses like utilities, oil and gas companies and mining; and infrastructure-centric companies like heavy transportation and pipeline operations.
Since the 2011 Hannover Fair, where the German government brought the idea to the forefront, the market has been coalescing around the cyber-physical construct that defines this fourth industrial revolution. The focus has been on design-to-build and smart connected supply chain aspects, and emphasized design and engineering opportunities. What has been missing from the Industry 4.0 discussion is the transformative impact on production assets.
Surveying more than 500 asset performance management (APM) practitioners, LNS Research sees an opportunity to elevate the asset reliability discussion to embody the full breadth of digital transformation. LNS refers to this APM evolutionary step as APM 4.0, to align with Industry 4.0.
Evolution of APM
Consider the first three generations of APM:
- Paper-based systems.
- Enterprise asset management (EAM) and computerized maintenance management systems (CMMS) displacing paper.
- The introduction of computer-based tools for condition-based maintenance (CBM), reliability-centered maintenance (RCM) and asset strategy definition, or classic specialized APM.
The fourth generation of APM leverages the full spectrum of digital tools and capabilities to redefine what APM is. The full range of APM 4.0 includes functionality spanning the process itself to the higher-level business systems, and includes reporting and transactional systems like CMMS/EAM, extends to predictive functionality associated with optimization, and encompasses a whole host of other tools and applications.
APM 4.0 becomes prescriptive
Of the “big five” of digital transformation—the Industrial Internet of Things (IIoT), cloud, mobility, Big Data and analytics—the one with the most growth is analytics. The other four have had their share of advancement as technology has become more affordable and powerful (thanks to Moore’s Law), but it is analytics that has the potential to radically transform APM as it evolves.
Historically, analytics was primarily used to describe or understand past behavior. With the advent of CBM and RCM, the focus shifted to predictive—describing what is likely to happen and when. For most organizations today, predictive analytics applied to reliability and maintenance is considered state-of-the-art. With the shift to APM 4.0, enterprises must go to the next level: prescriptive analytics.
Prescriptive analytics goes beyond predictive analytics by predicting what will happen and when, and providing guidance about options to deal with the issue. Simple predictive analytics uses process data to predict imminent failure in the next n hours and perhaps sends an alert for maintenance needed. With prescriptive maintenance, the analytics engine uses the same information plus data like production schedule, maintenance staff availability and forecasted scheduled downtime to offer options. One option might be to perform certain corrective maintenance at a specific time. Another might be to slow the process by 10%, fill scheduled production orders, and delay maintenance until the next scheduled downtime.
Of course, this capability requires access to far more data such as engineering design information, wear projections based on intended operating conditions, specific maintenance procedures, production and maintenance schedules, and more. This is why Big Data goes hand-in-hand with analytics in the APM 4.0 realm. Another element of prescriptive maintenance is the need for machine learning and artificial intelligence (AI) to act upon Big Data. The shortage of reliability and maintenance domain experts and data scientists makes this an imperative.
APM 4.0 is dynamic and visual
To make prescriptive maintenance a reality, there needs to be a digital twin, as a single piece of equipment plus the entire asset train such as a refinery or power plant. Design engineers create the digital twin as equipment is being built or the OEM delivers it when the plant is under construction. This means that OEMs must share design models with equipment users and design models must be dynamic models used to simulate performance, not just static construction details. Where equipment suppliers are unwilling to share details, companies should consider buying capacity from the machine manufacturer instead of the equipment. This approach allows the supplier to protect intellectual property but also places the onus of performance on them. Where end users do have a complete model of assets, they need to have a digital twin operating in parallel to real assets at all times. This posture allows jumping off into alternate paths to simulate potential outcomes for different operational scenarios.
To support the digital twin, APM 4.0 will drive adoption of augmented reality (AR) and virtual reality (VR) technologies. The mashup of visualization tools, simulation, modeling and real-time data acquisition becomes the “killer app” in the APM 4.0 world. By combining all these tools across a computing environment that combines cloud, edge or fog computing, and sharing the data across the global community of suppliers, sales, operations, customers and anyone else with the need to see what is happening, end users can maintain them with APM 4.0 capabilities, and get the most from Industry 4.0 plants.
You can’t buy APM 4.0
APM 4.0 isn’t a product or solution you buy any more than Industry 4.0 is a factory you build. APM 4.0 is, like Industry 4.0, an aspirational model and philosophy of how to maintain and operate a plant in context of Industry 4.0. No technology vendor or machine builder today can serve up a full suite of APM 4.0 capabilities. There are providers who, like in the IIoT realm, offer a platform with a large percentage of the capability. Similarly, there are APM 4.0 ecosystem suppliers that provide elements of functionality. Just as Industry 4.0 has taken five to seven years from concept to the first instantiation of plants that are 4.0 capable, LNS expects APM 4.0 to take a similar amount of time to evolve to support those plants.
You can't build smart products in dumb factories, and you certainly can't maintain a smart connected plant using yesterday’s APM technology. Industry 4.0 needs APM 4.0.
>>Dan Miklovic is a principal analyst with LNS Research, focused primarily on asset performance management and energy management, with collaborative coverage across manufacturing operations management (MOM), the Internet of Things, chemical, paper and packaging, metals and mining.
