Process plant personnel need perspective and insights into the key assets — boilers, heaters, burners, chillers, ovens, heat exchangers, furnaces, pumps and valves, etc. —used in their facilities. Many articles describe the technologies designed to provide those insights — those platforms, apps and plant automation devices that collectively help form the Industrial Internet of Things (IIoT). Though the array of options improves access to process insights, it can be a confusing world of technologies and strategies.
What is more, a key aspect often is overlooked — one that is the single most influential aspect in any organization. Collectively, the human element — workers, management and the company culture within which they operate — is a primary driver for continuous improvement. This article will discuss how plant personnel can be empowered and motivated to implement and embrace advanced data analytics and other IIoT-related technologies.
Technology is Just Part of the Equation
It is easier (and perhaps more entertaining) to focus on the wave of current innovations — what they can do — rather than looking at how to implement these technologies with existing plant personnel. Certainly, the revolution in sensors, data creation, wireless data collection and cloud-based data storage solutions has provided access to new insights and improved outcomes. The costs for the core system components have fallen to a fraction of what they were 10 or 20 years ago. As a result, they are ever more pervasive.
Advances also have been made in the acceptance and propagation of new analytics models for creating value from collected data. For example, consider how quickly industry is transitioning from a scheduled or preventive maintenance approach to predictive analytics for high value assets. Predictive analytics is just one of the analytics types familiar to most companies (figure 1). There are many more.
Many IIoT articles describe strategies and initiatives to help plants add IIoT technologies and insights. Industry 4.0, IIoT, smart manufacturing and digital transformation: No matter the term, the articles describe ways organizations can improve their operating results. (Within Process Heating, several recent articles do an excellent job of highlighting these points. For a few examples, turn to “Drying with IIoT and Cloud-Based Data Management,” October 2018, p. 27, or “Will IIoT Technologies Drive Plant Maintenance,” November 2018, p. 24, or find them on www.process-heating.com.) Through these and other articles, opportunities and strategies relevant to the IIoT innovations available to process plants are being addressed.
FIGURE 1. Predictive analytics follows a series of steps, a process greatly facilitated by using advanced analytics software.
But, in an environment with so much change and innovation, what has not changed is the human element. It is perhaps the most critical aspect of success in implementing these new technologies. As the famed management consultant Peter Drucker said, “Culture eats strategy for breakfast.” Plans do not always survive contact with reality — especially if they do not take company culture into account.
To address this challenge, organizations must answer three fundamental questions about their employees and their business priorities to balance technology innovation with objectives.
- What is the company hoping to accomplish by implementing IIoT technologies?
- Where does the company expect the IIoT competency experts reside: at the plant, at the vendor, or in both locations?
- How does the organization view its assets? How prepared is the company to rely on a data-centric approach to improving production insights?
Let’s look at each of these questions in turn.
Consider the Human Element
The first question to answer concerns the incentives and self-interest aspects of innovation in a plant environment. What is the company hoping to accomplish by implementing IIoT technologies?
If management’s goal for technology investment is to replace staff, then the reaction of those charged with implementing the solution will likely be negative, no matter how the initiative is presented to employees.
An example is an excessive focus on artificial intelligence and machine learning as a way to replace the expertise and experience of plant engineers with algorithms. This flies in the face of actual experience. Real-world implementations consistently demonstrate that advanced analytics cannot be applied effectively without ongoing interaction between engineers and other process experts with self-service software.
A more positive approach would therefore focus on how innovation can assist existing staff by accelerating their efforts. In some industries, the word “co-bot” is used for machines that assist and accelerate instead of the term robot, which suggests wholesale replacement of existing workers.
In the analytics environment, this has a software equivalent: self-service analytics offerings that accelerate the ability of engineers to find insights in large, distributed data sets (figure 2). By focusing on how various types of analytics — predictive, diagnostic (root cause) and descriptive (reporting) — can improve both the productivity of engineers and the plant, it is easy for staff members to see how innovation is a net positive for employees and organizations.
FIGURE 2. Analytic software is designed for self-service use by process engineers and experts, with no requirement for assistance from data scientists and other IT personnel.
Developing a Business Strategy and Applying Analytics
The second question looks at the business level and asks: What business is the plant in? In both the articles cited earlier, there are references to the connection between plant assets and the service providers or vendors that supplied the equipment. In a connected IIoT environment, this connection can be used for predictive analytics to improve asset uptime and the service experience, and for optimization of equipment operation.
But, this does not answer the question of what business the plant is engaged in. Is asset expertise a core competency for the plant in achieving production results? Or, is it something that could — or should — rightfully be associated with the asset vendor? Where do the experts reside: at the plant, at the vendor, or in both locations?
All that leads us to the third question: How does the organization look at its assets? Do they need to see them in the context of the process? How prepared is the company to rely on a data-centric approach to improving production insights?
In an environment where companies are buying flow instead of valves, or pressure instead of pumps, there is a new two-dimensional model where assets and support are orthogonal to the process. The result is that organizations will need to decide where their expertise should be focused, and where it is worth paying for external expertise that can be delivered by a constantly connected environment. And, of course, one needs to think about how this type of outsourcing will affect company culture (and bottom line).
To see how these business strategy questions might be answered, consider the following approach to analytics.
Within any organization, one must gauge the level of enthusiasm for a data-centric approach to improving production insights and, as one book title puts it, support for “competing on analytics.” Anyone with experience rolling out an analytics project will recognize the challenges when insights interact with — and sometimes conflict with — the gut feel of a senior manager. This does not mean results from analytics are always right. Leaving out key variables or considerations for more important priorities are common mistakes early in the process.
For example, data analytics can show what types of maintenance should be delayed to achieve production goals. The analytics might suggest the downside of such deferred maintenance is limited, but is it right? It depends on how thoroughly the analytics account for an ongoing interaction between engineers and other process experts with self-service software. This is analogous to taking a car in for service and being told an expensive repair — unrelated to what the car was brought in for — is needed. A novice mechanic might not have the expertise to know what can be deferred and for how long. So, the novice mechanic reports all repairs as equally important and urgent. By contrast, an expert mechanic who spots a problem with a car can use his or her knowledge to predict whether the owner must fix it now (due to safety or functional performance) or if the repair can wait, and for how long. Such advice may allow the car owner to save time and money by postponing an expenditure and scheduling the future repair work when it is more convenient.
Automated analytics may or may not be able to spot a maintenance issue. Either way, analytics cannot make intelligent decisions regarding tradeoffs between performing required maintenance now or later. These types of insights are perhaps the most valuable, however, because they combine speedy analysis with expertise and real-time priority setting.
The keys to accomplishing these insights and achieving broad buy in and participation are:
- Inclusion of all stakeholders early on and on an ongoing basis.
- Steadfast commitment to improving outcomes based on a shared definition of success and related key performance indicators.
Over time, this will result in increased confidence with the analytics the organization is using for decision making and will show the need for expert assistance.
With clarity on the answers to these questions, organizations can span the four necessary dimensions required to achieve improved performance:
- Technology innovations.
- Industry strategy.
- The personnel and organizational issues related to defining and accepting change.
In the following application examples, a self-service engineering effort for predictive analytics was used to achieve improved business results by spanning these four dimensions.
FIGURE 3. Analytic software was used to analyze declining performance of a feedwater heater and to make decisions regarding the timing of maintenance.
Power Plants Use Predictive Maintenance to Improve Performance
Using a model created with analytic software, a power plant identified declining performance in a specific asset. After analyzing the data, the team weighed the trade-off this decrease in asset performance against the market value of the plant’s output to decide when to perform maintenance. Such an approach takes the idea of preventive maintenance to another level. Rather than optimize for a process parameter or other metric, the power plant used real-time profitability as the priority outcome. In effect, real-time profitability became the setpoint for a control loop. This type of asset optimization is applicable to a wide range of equipment.
In another application, the power plant operators knew their feedwater heaters tended to foul and lose efficiency in a predictable manner. Because of insufficient instrumentation, however, they were never successful in their efforts to quantify the process so maintenance could be optimized. As a part of an analytics-drive solution, the power plant added the required instruments along with corresponding data collection and storage. Data analysis regarding the boiler’s heat rate provided the information necessary to determine what effect a cleaning effort had on efficiency, to the extent of determining its specific value (figure 3). Operators now optimize cleaning frequency based on the cost/benefit relationship.
This is the direction manufacturing is going: not to the asset or line level but to the outcome level of increased return on investment (ROI). A focus on outcome level ROI as the driver for manufacturing takes many shapes. It can mean delaying predictive maintenance that is known to be required so the plant can continue production. In other cases, it can mean taking the asset off line immediately to perform repairs. It all depends on the ROI analysis, driven by advanced analytics software.
In conclusion, process plant personnel must constantly keep up with the context and coverage of the opportunities available to them through innovations in technology, analytics and industrial strategies. They can ensure successful outcomes by considering the impact of these innovations on the people, culture and business priorities of their organizations.