BridgeGap Engineering Blog » Work Management

To Centralize or Not to Centralize . . .

GEARS — Thu, 30 Jul 2009 21:15:18 +0000

Lately, national politics seems dominated by discussions about the expansion of the federal government. Congressional proponents & opponents of such growth probably have a mix of benevolent and malevolent intentions – but evaluating their intent is misguided (as we all know what the road to hell is paved with!) When a new topic gets on the radar screen of congress, the first question that should be asked is: can the federal government realistically manage Idea X effectively? Or is this better handled by state & local communities (if not individuals)?

Similar ‘decision-rights’ questions should be asked inside our companies, especially in the field of RAM (Reliability & Maintenance). A corporate office is well-suited to address some topics; other times, it is just getting in the way. Where should the line be drawn?

One legitimate area for centralization is equipment naming conventions. Standardizing this across multiple plants can seem tedious and frustrating in the short-term – every plant already has a language for referring to equipment that is known and understood by the workers. But the enterprise also has a need to understand & address systemic problems. Doing so requires a common language. Without that consistency, every equipment name has to be time-consumingly re-translated into a common language, which is an extremely error-prone exercise. The result can be bad data, confused communication, and poor decisions that impact the entire enterprise.

However, there are other times when a corporate entity inappropriately exerts its influence. For example, it may be tempting to have a reliability engineer write preventive maintenance (PM) inspections for plant equipment and then upload that inspection into each plant’s Computerized Maintenance Management System (CMMS). But even if this is well-intended, it rarely works.

First of all, the most effective inspections are ones where the plant feels a sense of ownership. If the PM is dictated from afar, it may be greeted by the plant with irritation & passive/aggressive compliance. Secondly, PM frequencies will change over time, depending on equipment age. Unless the plants are exactly the same, the one-size-fits-all approach won’t always make sense. Third, inspections need to be logically clumped into common routes so they become schedulable events. This is very difficult to do from afar and may legitimately vary from plant to plant. Finally, if the plant cannot refine a PM without going through a maze of red tape, the inspection will be ignored or pencil-whipped. PM tweaks should be commonplace and easy to accomplish.

Here are some other examples of this dynamic:

Advantageous Centralization	Problematic Centralization
Requiring all plants to keep good refractory records. (Advantage: This will ensure that good data is available for continuous improvement)	Telling plants what kind of refractory they are required to use in each section of the kiln. (Problem: Kilns are different. Proper refractory choices only come from accurate, site-specific record-keeping)
Taking advantage of a corporation’s economy of scale by establishing global purchasing arrangements for commodity goods (Advantage: Increased profit margins through cost savings for the plant)	Enterprise-wide agreements for non-commodity services like machining, welding, or engineering. (Problem: Service quality may vary widely from community to community, even if those local service providers are owned by the same company)
Creating an expectation that all plants produce daily maintenance schedules. (Advantage: Resources are productively allocated when scheduled)	Creating the daily work schedule for the plant from the corporate level. (Problem: There is too much complexity at a plant for an individual to do this effectively from afar)
Building corporate support in specialized skills such as software programming, advanced failure analysis, or vibration spectrum analysis. (Advantage: Plants may not have those skill sets available locally. There may not be enough daily demand to warrant a full time employee with that expertise)	Failing to fully engage the plant in an evaluation of the skilled support that is provided. (Problem: The advanced skill set won’t add value if the plant has no buy-in for the recommendation)

The corporate office needs to strategically choose its battles and only push for centralizing the things that it can realistically manage effectively. If the wrong battles are chosen – or if too many battles are in progress – then the corporate office may lose the ability to manage the things it is obviously well-suited to address.

What do you think? When is centralization justified? When should decision-making be squarely in the hands of the plant?

Life’s How It Is, Not How It Ought To Be

GEARS — Wed, 15 Jul 2009 04:24:30 +0000

Have you ever heard the expression “Life’s how it is, not how it ought to be?”

An enterprise that wants to implement a standard reliability strategy across multiple sites must confront the implication of this statement.

Most mining enterprises do not have cookie-cutter plants or facilities. The locations will vary in age of equipment and workers. The quality of regional raw materials may be different. Plants near metropolitan centers may have high turnover while those in small towns may be staffed with “lifers.” Cultures can be different. Some plants are union free and historically have had good relations between hourly and management. Others have histories filled with mistrust and arbitrary rule enforcement.

Non-standard plant circumstances force the enterprise to find the right balance between centralization and decentralization. When do you standardize in the name of the enterprise and when must decision-making be left to local management?

Before some of these balances can be discovered, an enterprise must face the truth of “how it is.” Confronting the brutal facts of reality can be more difficult than it sounds.

If a company has spent millions of dollars on a Computerized Maintenance Management System (CMMS), it may not be pleasant to find out that most of the collected data is jumbled. If hundreds of thousands of dollars has been spent on vibration equipment & training, it may be painful to learn that the well-intentioned but part-time vibration specialist did not develop sophisticated time waveform analysis capabilities. If a loyal, friendly, experienced worker has been functioning as a planner for years, it’s hard to confront the fact his actual contribution is clerical in nature – and clerical contributions do not yield the productivity gains of planners who actually plan.

So – how does an enterprise face these brutal facts?

(1) Differentiate process from people. Remember that workers – especially long-time employees – are usually well intentioned. Organizations must consciously create work processes that focus on productivity & proactivity; these things do not naturally emerge without guidance and direction. A planner who focuses on non-productive behaviors is not necessarily a bad employee. The proper routines, tasks, and rules may not have been properly defined for him. When you view your organization through a process lens rather than a people lens, it is easier communicate without your employees feeling defensive or attacked.

(2) Hire a consulting firm. The reality is – the objective, third party gaze can often see things more clearly. Consultants aren’t smarter. They aren’t more knowledgeable about the operation of the plant. But if they are experienced, they are able to quickly zoom in on things that are indicative of a healthy reliability system. (Can the plant easily print a sorted list of Ready to Schedule work? Are the Preventive Maintenance tasks specifically defined?) As outsiders, consultants also have the luxury of remaining emotionally detached from existing relationship & histories between people & departments. This clinical disposition means they won’t be as easily swayed by charisma, earnestness, anger, or old wounds that won’t heal. Good consultants know they cannot add value to an enterprise if they are simply telling their clients what they may want to hear – or if they are ignoring the white elephant in the room. They know that a reliability initiative has to be built on a foundation of transparency and a clear assessment of current conditions.

(3) Establish norms around honest and respectful communication. Respect is important, as snippy exhortations about “ugly babies” and “garbage-in, garbage-out” data will get old real fast if the feedback has a tattle-tale quality. Stick to observations, not judgments about personalities. Accumulate multiple observations before drawing firm conclusions. Don’t just ask about problems that have occurred – ask about the frequency of those problems. Are parts/materials really never available when jobs are assigned, or did this only happen one time, three years ago? And never, ever retaliate against someone after you have given them permission to speak freely. If you do, you will never hear another honest observation again.

Reliability is a data driven field. It is a science. It has objective metrics. Although it may fail in the short-term (a vibration test may fail to identify a loose bolt) – over time, consistent application of reliability concepts produce huge dollar savings. Sycophantic relationships will destroy reliability. Transparent, data-driven decision making must be the norm; otherwise, reliability is just a meaningless label to smack onto old habits. But before an enterprise can begin its reliability journey, it must have an honest assessment of where it is at.

What are other ways an organization can get a good assessment of its current reliability situation?