Behind the dashboard

Who hasn’t fantasized about being in control – able to see what people are doing, how equipment is performing, where problems are holding you back – in real time, from a position of command, comfort and convenience? Well, the bridge of the starship Enterprise still exists only in the imagination, but visual systems that allow technicians, engineers and managers to quickly ascertain the status of industrial equipment are very real, and on their way to a plant near you.

Dashboards accept data from disparate sources, make agreed-upon calculations and present the resulting information as individually customizable, easily scanned displays.

Data may be drawn from finance, maintenance, manufacturing, supply chain management, workforce management, customer relationship management or other sources. Indicators are typically color-coded green, yellow or red to indicate status at a glance, and information may be many layers deep, allowing users to drill down from a trouble indicator to details about the plant, line, machine, component or individual employee involved.
Only a few years ago, the concept of displaying in detail the condition of a single complex plant, let alone the many facilities of a global enterprise, was as farfetched as Spock’s ears. I distinctly remember the response of a dashboard vendor spokesman when asked, after his glorious presentation of the company’s package, “What does it take to get the information into the system?” He said, “That’s not our problem. We do faucets, not plumbing.”

Today, thanks to developments ranging from Y2K updates of ERP software through MIMOSA, OPC-UA and S95 to HTML and Web Services, effective dashboard systems can be relatively straightforward to plumb. But why might you want one, what does it take to do it right, and is it worth it?

View related content on PlantServices.com

Focus on information

In the realm of industrial asset management, the most interesting dashboard applications are for condition monitoring, workforce management and overall equipment effectiveness (OEE). Their power lies in bringing the same information to multiple departments and individuals in the format they find most useful for doing their jobs.

“When we assess plants, we find a hodgepodge of information that does not allow a complete picture of plant performance,” says Erich Scheller, principal consultant, Life Cycle Engineering (LCE, www.lce.com). “Separation and segregation of departments is a hindrance to complete information and display, both within a site and over multiple sites.”

Although the technological walls have been broken between automation, operations and maintenance, “Business practices that would enable companies to truly benefit from this newfound freedom have been slow to change,” says Ulf Stern, co-founder, IFS (www.ifsworld.com). Breaching the walls can allow common access to floods of data, but typically doesn’t improve communication.

“The results of human/computer interaction (HCI) research suggest that visualization tools are essential for manipulating and interpreting large quantities of data,” says Stern.

“Timelines and tree maps are especially effective in allowing plant operators to distill usable information from hundreds of thousands of operational metrics.”

For example, Air Liquide monitors and services gas generation, storage and delivery equipment at thousands of sites. “Our data showed discontinuity between action items identified and action items completed, which we attribute to lack of flow of information,” says Steve Busick, manager, maintenance and reliability national programs, Air Liquide (www.airliquide.com). “We felt it was important to stop moving information from platform to platform, because it might not get from one platform to the next.”

The company piloted a new dashboard system from Rockwell Automation during the first half of this year, and started using it full-time in June at 113 sites. In the old system, information had to make it from a route report through engineer review, a summary report and recommendation to the reliability engineers, an action plan by the reliability engineers, work order generation and work order status reports. “Now, the only stop is to get the information into Maximo, which is not yet fully integrated,” says Busick.

Air Liquide is rolling the system out to more service sites and adding other tools and services to report performance. “One of the big benefits we see is the engineers don’t have to spend as much of their valuable time messing with the system, so they are doing more engineering. The system works for them instead of them working for the system,” Busick adds. “Also, the tighter integration means information never comes to a halt. The flow of requests is automated, and critical issues are focused and kept in the forefront.”

It’s not uncommon for plants to keep condition monitoring data such as vibration, oil analysis and thermography readings in separate systems. “There was no easy way for them to use all three to make decisions,” says Ralph Aleseo, business manager for integrated condition monitoring, Rockwell Automation (www.ra.rockwell.com). “We designed a Web tool to make it possible to look at all three, along with maintenance histories, similar equipment all over the plant, and warranty information and experience. It covers multiple plants with drill-down, is integrated with the CMMS for work orders out and complete, and is tied back to ROI.”

What do you want to see?

Dashboards range from relatively small-scale systems for monitoring the condition of a few pieces of critical equipment or managing a plant’s maintenance workforce to multiple-facility, global systems for tracking minute-by-minute plant performance. Regardless of the subject or size, all depend on the same keys for success: enabling technology, the workflow process, and cultural change.

“We call them process, technology and people,” says Eric Houston, vice-president, Asset Management Services, SKF Reliability Systems (www.skf.com/reliability). “The technology is the easy part. There are a variety of tools on the market – SAP®, MS Office, Internet browser - all can work. Use what you have. Your business scenario determines the technology.” The more difficult concerns are establishing an effective workflow process and implementing culture change. “Especially culture change,” says Houston.

Experts are of a single voice when describing the importance of understanding goals of the organization, including the stakeholders, and obtaining buy-in before deciding what information will be gathered and how it will be processed before it’s made available through dashboard displays. The company’s vision, mission and business goals must be understood. Goals must be SMART (specific, measurable, attainable, realistic and tangible), as must be KPIs (specific, measurable, agreed-upon, relevant and timed).

“When we work with a customer to set up KPIs, we start with the vision and goals to end up with a balanced scorecard that takes into account production, maintenance, safety, energy, efficiency, and taking care of the environment – all the factors,” says Houston. “We do an assessment and gap analysis, and use our industry-specific benchmark data to determine which of four levels the company is on for 40 KPIs.”

Stakeholder involvement makes or breaks the project. “Reliability improvement programs tend to seem like flavor-of-the-month,” warns Brian Maguire, reliability expert, Ivara (www.ivara.com). “You must make collecting the data a part of people’s jobs, so be sure they can see how they’ll benefit from it. If the information supports the individual’s objectives, they’ll support it.”

The dashboard gives high-level, processed information. “It’s critical to understand where the data is coming from, how it’s collected, and how the calculations are being made,” says Glenn Schultz, director, asset management business for software, Rockwell Automation. “If these are not sold and signed off, the results can be useless.” Managers will ask for information that requires additional data. “If users don’t understand the context, they won’t want to provide it,” says Maquire. Involve plant people in defining what information needs to be obtained. Aligning management and the plant gives meaningful information, and people are more willing to collect it.

For higher-level systems, the challenge is to define key performance indicators (KPIs) that are meaningful both for the business and for the individual using the dashboard. “The approach needs to be role-based, with an integrated set of KPIs that will support improved communication across all the key groups – engineering, maintenance and operations,” says Neil Cooper, general manager, Invensys’ Avantis unit (www.avantis.net). “To fully understand the impact on the business, the data should also include real-time accounting information.”

Once the required KPIs are identified, the bulk of the effort is determining if the information is available and in what form, then setting up the necessary infrastructure to extract, transform and load the data to the business intelligence (BI) solution. “The majority of the BI solutions provide powerful tools to support the data management processes,” Cooper says.

“Properly applied, this role-based and integrated set of KPIs allows the management team to understand the interrelationships of all the groups as well as the economic impact of decisions,” he says. “It provides a broader and more complete context for decision-making in all groups, which is critical as organizations move from a traditional asset management approach to asset performance management.”

Looking good

First and foremost, make sure the dashboard system shows the end-user information such that they can take action or make actionable decisions based on the information (Figure 1). “All other information may make nice pictures, but serve no business purpose and will drag down the ROI,” says Frank Vanderham, manager, equipment and process monitoring solutions, Matrikon (www.matrikon.com).

Figure 1: A dashboard system should allow users to readily take (split infinitive) action or make actionable decisions based on the displayed information.

Dashboards too often become just graphical displays of numerical data -- very technical and not very informational, and not tied to KPIs. “It’s like a car with a voltmeter, but no charging system failure lamp,” says Dan Bradley, president, SKF Reliability Systems.

“People need clear indicators that tell them if their initiatives are taking place. It’s all about changing organizational behavior.”

Avoid presenting raw data. Instead, show information in the context of the operation’s physical assets or logical units. “This requires assembly of different data sources (i.e. lab, CMMS, historian) into asset-centric units (i.e. compressor, Upgrading Division),” says Vanderham. “It also requires roll-ups, for example, from time-based to shift-based.”

Present information seamlessly on a single dashboard without requiring the user to switch among applications.

Present auditable information. “It needs to be resilient against the inevitable communication dropouts to field instruments and occasional network problems between the integrated systems,” Vanderham adds. “The ultimate aim is to have the right people take the right action based on the information on the dashboard.”

Presentation should be visual and easy to use (Figure 2). “For example, use a photograph of the plant that shows where a problem is and lets you drill down to indicators with just a couple of clicks,” says Maguire.

Figure 2: Start with an intuitive overview and allow users to drill down to details with a few clicks.

Use leading indicators that are as real-time as possible so you can do something about it today – talk to the guy on the floor, get more training, allocate more resources -- and do a better job (see “Get out of the dark,” August 2006, page 40).

The dashboard should be configurable so users and management can define the information they need. “We provide standards and definitions, but they customize,” Maguire says. Different plant managers want to see different things based on their experience and how they perceive their plant. “We have one manager who had been named responsible for an accident due to a non-functioning protective device. He wants to be able to see the status of all the protective devices so he won’t go to jail.”

Consider thin-client Web-based technology so dashboards can be accessed from anywhere within the corporation using PDAs, Blackberries, Pocket PCs, etc.
Include e-mail, phones and pagers. “Much of the workforce is disconnected,” says Schultz.

“Rarely does a maintenance guy have time to sit and look at a dashboard, so it’s important to alert users by e-mail, text message or cell phone to look at the dashboard.”

Determine setpoints carefully. Setting alarms on indicators simplistically can lead to problems with dashboards that don’t catch problems early enough, or constantly cry wolf until no one pays them any attention. SKF’s decision-support system solves that problem by applying the company’s experience to condition-monitoring, process and inspection data. Its alarm setpoints are lower than in a typical condition-monitoring system. “Several low-level alarms may mean something is going on, so you see problems at levels where, if you set them as condition-monitoring or process alarms, you’d have alarms all the time. It’s the knowledge on how these readings interrelate that provides the advantage,” says Scott Brady, director of product marketing, SKF Reliability Systems.

“The challenge for most applications is the knowledge it takes to do the diagnostics,” Brady says. “In the ’90s, that knowledge was limited. Now, we’ve spent many man-years developing it, and the average guy can implement decision support. They can start with our knowledge base and, over time, add specifics for their plant.”

When the decision-support system detects a failure, the user sees it, sees the criticality, and can create a work order. Some users have integrated it with the CMMS so they can generate a work order within the decision-support system. “But while decision support is a proven approach with demonstrated payback, the return on integrating it with the CMMS is grayer, and not all CMMS companies cooperate,” Brady says. “It’s practical where they have written interfaces for us, such as SAP and MRO Software. The challenge is adoption by CMMS suppliers – some do it because they see it as the next avenue of adding value, but others will not return our calls. End users need to demand it.”

Shortsighted IT?

As mentioned above, piping up a dashboard system is now well within the capabilities of most plant’s IT departments. “The technical infrastructure is fairly simple, generally requiring a central server where the data can be consolidated,” says Cooper. “Access is generally provided through browser-based tools and is delivered over the corporate intranet or over a secure Internet connection.”

But like any IT project, it’s easy to get into trouble, and the best way to stay out of it is to involve IT from the outset. “You must clearly articulate the vision, and rely on IT’s resources to make it happen,” says George Hofer, Air Liquide corporate program manager, Rockwell Automation. “Most organization’s IT departments are intimately familiar with their systems architecture and how they communicate – the databases, infrastructure, connections, firewall issues, etc. Make sure any outside development is done in coordination with your IT people.”

You say you have no IT infrastructure for this project? You’re not alone. Many organizations are using separate, manual systems. “They can supply Excel-type spreadsheet information to someone who puts it all together,” says Scheller. “Putting it all together in a central database is too big a project, so very few companies have done it, but an increasing number are finding a way to put it together on a daily basis.”

LCE helps its clients install dashboards as part of its program to improve reliability. It recommends a benefit-tracking system with KPIs for five levels from plant-floor to company-wide, teams to provide information on maintenance costs and asset utilization, and an individual responsible for putting the information out on a daily, weekly and monthly basis.

“An LCE principal is on-site at the beginning to act as a conduit and coach, typically for four to six weeks,” says Scheller, “Then a team leader and subject matter experts from LCE continue to mentor and coach. It’s a well-thought-out, methodical process, and we won’t go away until they get there.

“The dashboard’s role in culture change depends on top-level management having the courage to go through with it. Our clients save millions of dollars a year, not by head-cutting but by changing the way business is done.”

If your headcount leaves little to devote to a dashboard implementation, there’s no shortage of vendors and consultants ready to help. For example, Advanced Automation’s dashboard is populated using the company’s CMMS, data automatically collected from the plant, and manual input. The system is maintained by Advanced Automation.

Bringing a new client up on the system is a two-stage process. “We assess their plant and then recommend the level of service that will meet their current and future needs,” says Robert Vogel, director, support services, Advanced Automation (www.advancedautomation.com). Service level needs vary according to the current plant efficiency level. “We offer ‘step-down’ service contracts as many clients need to go through a stabilization period first, before we can begin to maintain and improve their processes. This period usually requires daily on-site support, tapered down as the plant operation becomes more consistent.”

Many plants do not have a good grasp of the current structure (system details, parts inventory, documentation, file and program storage, etc.) within their plant, so the first step is an on-site assessment. “This provides us with the ability to get our arms around the client’s plant structure and architecture,” Vogel says. “Once all of the information is gathered, it is loaded into our maintenance management system.”

Preventive and predictive maintenance schedules are established according to asset; service issues (break/fix) and engineering requests are tracked and reported by asset; and current inventory levels are logged and tracked by asset. “Depending on the size of the facility, from assessment to the client being able to display data on the dashboard, the process could take a week to a couple of months,” Vogel says.

20/20 vision pays off

A dashboard is a good tool for keeping resources focused and giving management proof of results. You can capture failure modes, consequences and costs, and show metrics that capture the ongoing value of avoided failures to show management how proactive mechanisms avoid those costs. “You track the cost of monitoring versus the cost of failure, and show, for example, that you spent $1,600 to save $24,000,” says Maguire.

You can also see the costs of poor management practices. “One company suspended predictive activities to cover summer vacations, which led to a spike in breakdowns,” Maguire says. “If management has visibility into the consequences of a reallocation like that, it can understand the costs and make the right decision to maximize returns.”

Some companies have adopted a system by which they put in a plan for the year with targets established for the plant or plants. Everyone has the same definitions, and they measure month-by-month. If, for example, a plant’s work control is red for a critical period, they put in a corrective action team. The problem is documented and actions are taken.

“It’s a very well designed process based on company-specific targets and indicators, and it links back to all the departments so everyone has skin in the game and works to a common set of performance indicators,” says Scheller. “This leads to culture change and the view of maintenance and reliability as a business.”

Other key benefits include:

Make better decisions on a daily (or perhaps even hourly) basis. Dashboards provide maintenance supervisors and technicians with prioritization on where to focus maintenance activities and expenses. “The primary benefits for all groups come from the timeliness and visibility of key information, and the improved decision-making that result,” says Cooper

Quantify the maintenance department’s output in terms of mean-time-between-failures (MTBF), overall equipment effectiveness (OEE), reliability percent, availability percent, etc.

Show how well maintenance is performing compared to targets, using metrics such as PM compliance, bad actors (equipment requiring the most maintenance hours/interventions), and backlog.

Focus your meetings. Daily meetings between maintenance and operations typically involve a lot of finger-pointing about yesterday’s problems. A dashboard can show the problems and help them talk about plans, which is much more constructive.
Show the results of work by reliability engineers. Their critical contributions to plant performance are too often poorly understood and unappreciated.

Put your house in order. Dashboards help maintenance look more organized and professional. “Production has always been able to do this, and now maintenance can, too,” says Maquire. Instead of having information sitting in many databases – vibration, thermography, etc. – it can all be in one place.

This is also good for regulatory audits. “For example, a nuclear plant undergoing an Institute of Nuclear Power Operations (INPO) audit can present it all, and it’s all kept online,” Maguire says. “The audit and review goes more smoothly and it shows a strong managerial approach. If you’re in a maintenance management role, this ability can help your career.”

Big Brother is watching

High visibility and easy, plant-wide access to information can help quantify the value of your work and make it obvious to upper-level management, which can be a mixed blessing.

“The first plant floor objection we usually have to overcome is the perceived threat by the client’s engineering and technical staff that we are there to replace them,” says Vogel.

“This is not the case. In most instances, we are there as the expert, tasked with getting more out of what they have, using technology instead of additional manpower or upgrading equipment.” He says in almost all cases, once the perception of threat has subsided, the engineers and technical staff find the file and document storage structure easy to use, and the dashboard indispensable in support of the plant.

Dashboards provide higher-level management with maintenance and operational information in the broader context of productivity. “Without dashboards, it’s not uncommon to push higher demand on the asset, which causes extraordinary wear and tear and eventual equipment failure,” says Vanderham. The dashboard brings maintenance and asset health into the decision-making process. A better decision can then be made to use lower demand with longer asset health, which yields higher overall productivity.

Any time you are able to provide real-time information surrounding operational efficiency and maintenance, you take the blinders off of management. The goal is to provide them with the proper tools they need to better manage their plant resources, parts inventories, maintenance schedules and improve line output.

“It’s kind of a double-edged sword,” says Air Liquide’s Busick. “The information is very good – it shows the rigor, the engineering discipline, the significant amount of work behind the predictive finds, and the dollars spent per save. It helps everyone see the value of knowing the health of the equipment, but it can expose upper-level management to a level of detail they’re not familiar with. You want to make sure your management understands what they’re looking at.”

It’s important that the dashboard speak their language. “Upper-level management is only looking at financials — production quantity, quality and cost. We have to make the connection,” says Bradley. For example, a company might have a corporate initiative to reduce energy consumption by 5%. Maintenance links motor cleaning and precision alignment to that initiative. “We need to ask for and get a list of initiatives and design our KPIs to reflect them,” he says.

In the end, it’s all about proving ROI. When maintenance can document avoided costs, see failure causes and track downtime not directly related to maintenance, costs can be assigned where they belong. “We made sure the system could do this,” says Aleseo. “We don’t ever want to lose a project because we couldn’t articulate the savings.”

--------------------------------------------------------------------------------