Clearing up concerns about asset condition monitoring systems
My primary job is helping customers solve problems using the latest technology. But I also spend quite a bit of time making presentations at various industry conferences talking about that technology. In the last few months, I’ve traveled coast-to-coast ‒ making a presentation in Washington DC one week and another in San Francisco a few weeks later.
My topic recently has been how to avoid power outages and related shutdowns in data centers. As you can imagine, that’s a topic that gets a lot of interest. One of the most effective ways to prevent outages is using what’s known as an Asset Condition Monitoring System (ACMS).
What an ACMS basically does is tie together a center’s sensor-enabled devices (transformers, switchgear, battery backups, etc.), data collection, and decision-making systems. With all of these assets connected, much of the diagnostic and repair activity is automated and improved. When human intervention is required, operators and service people have all the information they need to identify and prevent potential power problems.
Actually, preventing outages is only one thing these systems accomplish. The ACMS will also increase asset utilization, improve equipment performance and productivity, and enhance safety. When attendees hear about all the tremendous things an ACMS can do for them, they’re typically highly excited. But in the next breath they will tell me that it’s unlikely they’ll pursue the idea of an ACMS for their data center.
Why is that? The biggest reason is the same one that has truck drivers shaking in their boots right now; technology will put them out of a job. Self-driving vehicles are here. A friend who visited Pittsburgh recently saw at least a dozen self-driving Uber cars, and Anheuser-Busch InBev just made the first-ever delivery (45,000 cans of Budweiser) via a self-driving truck.
In data centers, managers are concerned that an ACMS will automate monitoring and maintenance to the extent that their jobs will be in jeopardy. This is certainly a valid concern. Dark data centers, administered remotely with no regular human intervention, are a reality. Robotic systems are available that will pull and repair or replace troublesome servers. As in most activities, automation provides more effective and efficient operations with tremendous and undeniable benefits.
Still, there will always be a need for humans in management roles at data centers. Their role, however, will shift from the hands-on maintenance tasks to higher-level planning, budgeting, and operations-management responsibilities.
Another reason I hear from people who express doubts about implementing an ACMS is their fear of the complexity involved. While there is some complexity, implementation continues to become simpler as the technology improves.
At the asset-level, many – possibly most – new power-control devices are available with the needed sensors from the OEM, either as a standard feature or an optional add-on. These sensors communicate via standard protocols, nearly eliminating the need for custom software or hardware interfaces. This enables straightforward connections to control systems. The control systems, too, continue to improve in functionality and usability.
Of course, implementing an ACMS will require the guidance and support of a knowledgeable partner. In the last few conferences where I presented, I was proud to point out the recent Navigant Research ratings* that named ABB as one of the two leaders in asset-performance management, which includes ACMS. This rating related specifically to our work with utilities, but we are increasing the application of our knowledge to power systems used in manufacturing, process industries, and ‒ of course ‒ data centers. Electric energy and its distribution and control, is largely the same regardless of the application.
In all of these markets, when facility owners or operators deploy an ACMS, they begin to realize the benefits of a new approach to maintenance. Instead of relying on routine, scheduled maintenance or a break-fix approach, they now have asset-health data and the supporting algorithms to identify potential issues, typically before they result in an outage. That means maintenance crews can focus their attention on the assets that actually need attention rather than spending their time checking on equipment that’s working just fine.
The return on investment for an ACMS deployment, at least for now, isn’t there at smaller data centers. A better option for them may be simply to add redundancy of their key power-related assets. The ACMS solution is focused on enterprise-level centers, where the scale of operations and the cost of an outage warrant the investment in ensuring power at the six nines reliability level. Delta Airlines could have benefited greatly from an ACMS. A simple sensor would likely have detected the early temperature rise and warned them to have a look, probably avoiding the crippling system outage they experienced.
I will be back on the road again in a few weeks, attending another conference and continuing to enlighten attendees about the benefits they can realize from an ACMS. I expect that most will continue to be excited by the incredible benefits these systems provide. I hope that more of them will agree that the long-term benefits outweigh the short-term concerns that accompany the introduction of almost all new technology, including ACMS.
*http://www.navigantresearch.com/research/navigant-research-leaderboard-r...