American Electric Power Streamlines the Acquisition and Processing of Earned-Value Data

Using a combination of existing toolsets and domain expertise, the company successfully updated its Earned Value methodology in a critical system switch-over.

The fundamental aspects of all great projects, stated in an oversimplified way, are cost, scope, schedule, and risk. All must be tracked and managed in the most cost-effective way possible. But what do you do when the system being used introduces significant cost and risk itself? This was the dilemma facing the Generation Project Controls and Construction group at American Electric Power, one of the Midwest’s great power providers.

Baselining

Most systems begin with the best of intentions, and often the best of justifications. Fundamentally, you can only use the tools at hand, though, and what seems like the most effective toolset at a project’s inception will, inevitably, become obsolete and inefficient over time. Earned Value Management was, and remains, an important concept for AEP Project Controls and Construction (PC&C) and a system was commissioned to automate the reporting of Earned Value (EV). The system was built by an individual in a consultant role who possessed EV and programing experience. The system was deployed and provided AEP PC&C with its initial automated and consistent EV reporting tool.

Employing the technology available at the time, the system was developed using an Excel-based methodology that met the requirements desired by AEP PC&C. The system had some features that provided key metrics the company needed, but it also posed some risks to AEP PC&C First, it required a fairly advanced Excel skill set. Second, it required a great deal of direct domain expertise, just to use it. Third, it was very labor-intensive, with multiple manual “touchpoints” in the process. Lastly, it was highly dependent on specific contractors and individuals; if someone was unable to perform their part in the process because of a vacation or illness, a lot of anxiety ensued. But perhaps the most significant drawback was the single point of failure: Only one individual was capable of maintaining critical parts of the system. This introduced a relatively high degree of risk—and cost.

Risk Management

The irony of a high degree of risk imposed by a system that is supposed to help reduce it was not lost on AEP’s upper management. They decided to mitigate this risk by rearchitecting the system using a combination of off-the shelf tools from well-established entities. However, since “nothing lasts forever,” it was also important to avoid putting the company in the same jeopardy they were already in. The system would need to contain components that could be exchangeable. And changing it out would need to avoid disrupting operations as much as possible. Critically, the data generated by the system would need to align the company’s contractor payments with actual project status.

The Big Switch

Collecting, staging, and aggregating project Earned Value (EV) data is, by its very nature, somewhat complicated. Finding a way to swap one system with another was going to be a difficult undertaking, no doubt, especially when the existing system was working and producing highly-tuned, accurate results. Any new system would need to produce at least what the old system did, but in a more automated, easier-to-maintain way. AEP knew that a detailed, up-front system analysis would need to take place and a thorough requirements document would have to be produced. A team was assembled and work began.

When Days Become Nights and Requirements Documents Become Novels

Naturally, the most difficult part of any project is assembling the right team members, people with specific expertise in various aspects. Often, team members come from diverse backgrounds. Getting them to work well together is a challenge, even for the best of leaders. With all committed to the success of the project, they will want to make sure their input is included. “It’s vital to have a single point of authority over what is a real requirement. You have to be able to distinguish between ‘must’ and ‘should,’ balancing functionality with cost,” says Kelly Ruppert, V.P. of Technology Delivery at Knowledge Relay (KR). The users and administrators of the existing system held an interest in keeping all of the existing functionality. In the end, some of that functionality was deemed unnecessary to implement in the new system.

Of course, regardless of risk and cost, the legacy system was successfully providing many useful features. This drove many debates about matching real requirements to functionality. “It’s a difficult balance,” says Steve Mauss, KR’s CEO. “Many things are nice to have in an EV system but they don’t necessarily align with core requirements. They can be added in a future release, if desired, or eliminated altogether.”

The Foundation is Laid

One of the project goals was to use as much off-the-shelf technology as possible, tools that are supported by well-established companies. Of course, Excel is supported by Microsoft, but it is a tool used for building solutions, not the solution itself, anymore than a hammer is a wall. Since many of the advanced features of Excel are beyond the understanding of most users, using it as the basis of a complex EV system would be akin to expecting users to know how an internal combustion engine works in order to drive a car. The system would need to be much simpler to users and be, to the extent possible, automated and available at the enterprise level. That’s where the foundational toolset comes into play.

Nearly every large enterprise uses some form of large-scale Database Management System (DBMS), and AEP is no exception. Most of their systems rely on Oracle™, including Oracle’s own Primavera P6 Enterprise Project and Portfolio Management (EPPM) system. Even the company’s homegrown cost management system (PMM) utilizes an Oracle database. When bringing in costs from multiple vendors, it’s important to have a clear understanding of the DBMS back-end. This understanding contributes directly to knowing how to automate the data migration effort as costs are compiled, how to spot anomalies, and how to perform accurate calculations. AEP decided to use Knowledge Relay’s DIS™ for this portion, since KR is a recognized data migration expert in the energy market space (used by over 70% of nuclear power producers). Once the data is moving correctly, the ability to display it needs to be in place. It, too, has to be easy to operate. Most consumers of the information have to be able to see what’s important to them with a single click. The best outcome is a simple portal that can be tailored to specific users and groups. For this portion, AEP decided to use a combination of Knowledge Relay’s Information Visualizer™ reporting toolset and Dundas BI™ dashboards. These tools automatically create and post the visuals needed by the team to a common portal with essentially no user intervention. A schedule is created and they just run. And while the display tools chosen are considered the “best fit” now, it is important to note that any effective system should remain “tool agnostic” to the highest degree possible. If AEP ever needs to swap reporting and display methods, their new system will continue to function as designed.

It’s Just a Matter of Time…and Risk…and Cost

With the proper foundation laid, the developers and consultants went to work. Multiple meetings were held throughout the project to ensure that the requirements document was correctly written and that the metrics being delivered were accurate. The development staff worked out the data modeling and created a “Data Depot,” an integral step to stage data for validation and performing the needed calculations.

With the Data Depot in place, the developers were able to begin testing of the visuals. “It’s almost impossible to get the visuals perfectly aligned with the data until the data is correct,” says Jeff Muus, Knowledge Relay’s Technology Delivery Supervisor. “There are a number of nuances that have to be considered, such as platform and display mechanism. It’s so important to have a toolset that is flexible enough to adapt to whatever comes your way. KR’s reports and Dundas BI dashboards had just the right combination of power and flexibility, without over-complicating delivery.”

By assembling the right tools and absorbing the complexity, AEP was able to create an Earned Value system that is automated and maintainable with a lower risk and significantly reduced cost. The resulting information is posted to a web-based portal, greatly simplifying access to key decision points and payment schedules. The company can now more easily align their project costs with what is showing in their accounting system.

Lessons Learned

During the Close-out phase (5), the team took some time to evaluate what made this project successful. Here are their findings:

• A robust and thorough requirements document is the foundation of every great project. This is where the AEP/KR team spent nearly a third of the time.
• Domain expertise, including knowledge of source and destination systems, is vital. The old idea that there is always a simple solution to a complex problem must be abandoned while, at the same time, the complexity of any system must be encapsulated and automated to the extent possible. Humans must be freed up to invent and create.
• Using already-available toolsets in different ways to accomplish new things has huge benefits. Existing systems are likely already “battle-tested” and supported, and can save a lot of development effort.
• Aligning progress payments to contractors with actual, real project status reduces risk and costs to a workable level.
• Never get caught being tied down to a single consultant to develop, manage, and maintain a mission-critical system. Try to use a well-established company with a solid industry reputation.
• A strong Project Manager can make all the difference. It’s important to have someone who consistently pushes toward the finishing line, making sure the project stays on schedule and budget. There needs to be a reliable single decision point, after all points of view have been aired, to determine the way forward.