Downtime avoidance: Proactive support services seek to eliminate diagnostic delays

Predictive monitoring and automated problem diagnoses and resolution can give quick resolution of problems. Invensys is using software from NextNine to keep its customers' plants running.

By Control Engineering Staff July 10, 2008

Manufacturing automation and control systems are increasingly complex, and when plant engineers have a problem, they are asking support service providers to do more. That means vendor support organizations must handle more calls, more quickly and provide better answers to hardware and software problems. While some support service organizations are addressing these pressures by creating knowledgebases for their service engineers to use, others are establishing online communities to enable more peer support. The majority, however, are seeking methods of proactive support, which involves predictive monitoring of equipment and automated problem diagnoses—and even resolution—before a service call is made. The result is fewer emergency service calls, quicker resolution of problems when they do occur, and significantly less downtime for essential user processes. Invensys is using support center software from NextNine to provide such benefits to its customers.

l while plant processes are slowed or shut down.

However, a Service and Support Professional Association (SSPA) News poll from January 2008 said 42% of service organizations planned to invest in proactive support technologies in 2008.

What is proactive support? “For a support organization, it means having your best supportof problems. It has healthcare, telecom and manufacturing company support organizations as customers.

NextNine’s Virtual Support Engineer is agentless Java software that securely resides at the customer site. It’s an automated service that does data collection, real time diagnosis, and command execution. It can generate proactive alerts based on equipment or process symptoms and securely send them to remote support engineers for additional action.

Gerry Murphy, director of services portfolio for Invensys, says his company has been offering remote support for 16 years, primarily for distributed control systems, based on software they built themselves. The service is called Remote Watch (and used to be called Fox Watch.) “We had the ability for one of our experts to log in once a problem happened and diagnose the problem. That was beneficial to us, because it helped us understand the problems more quickly. We could decide whether to send a service engineer to help, but it was still too late to head off the larger degradation of process. Our goal for last five years has been to be more proactive, to find problems before customers find them.”

Finding problems before they occur has benefits to both users and automation vendors.

“Between their [users’] loss and our need to put a lot of resources into place, often under duress, it cost a lot of money to be reactive,” says Murphy. “We had built our own solution. It met our needs, but it was inflicting a little pain on the system. So the question became, how could we be proactive without taking up too much resource in the system we sold. We thought [NextNine’s] solution was the most robust at the time and did what we needed.”

Invensys is building up a knowledgebase of if-then scenarios for problem diagnosis, but the NextNine software provides “a history that we can trend. We now have the ability to historize the data, so we can quickly see when it’s happening and predict what will happen next. We have a historian that we embedded, but now we do the data acquisition with NextNine,” says Murphy.

Some issues are basic: Disk space is filling. There are network issues with lots of collisions. Others are more rare: A batch application launches and certain conditions occur that cause problems elsewhere. “Before, we could look in and see what was happening right then, but we couldn’t look in the past,” he says.

Knowledge bases are product-specific, but support engineers also have to understand what’s going on at the site. The challenge from Invensys’ point of view is that the products are so complex and there are so many variables, that it’s hard to have enough “experts.”ia or on platforms in the ocean somewhere—and almost 100 are using the new NextNine solution, he says.

“It’s part of our service agreement solution with DCS, and we’re looking at rolling it beyond: to safety systems and advanced applications,” says Murphy.

Invensys deploys the service through a server at the customer’s site, and has to go through manufacturing IT groups to make the connection. Security is of course a big issue, says Murphy, and “sometimes we’ll be the mediators in the discussion with IT.sed location. Now, we’re going through the front door, through the internet, and we have other security measures in place.”

Murphy says, “The major oil companies have agreed to the solution, because we’ve been able to solve their problems quicker with it.” And, because the software is deployed on a local server, users have access to the same tools and data that Invensys engineers do. That may give them an added sense of security, but it doesn’t necessarily mean they use the tools themselves. “They have less time than ever to manage vendor systems,” says Murphy.

-Edited by Renee Robbins, senior editor, renee.robbins@reedbusiness.com ,
Control Engineering Information Control eNewsletter
Register here and scroll down to select your choice of free eNewsletters .