Siemens Uses AWS to Reduce Power Plant Alerts by 90%

It's never been easier for power plants to monitor their equipment and operations, but the ease of deploying thousands of sensors results in a challenging volume of sensor data. That’s a challenge Siemens, the 170-year-old global technology leader, set out to answer. The solution Siemens Gas and Power built on AWS has decreased the volume of alerts for its power plant customers by 90 percent.

"A modern power plant control system receives about 5,000 alerts about component failures or out-of-tolerance process conditions a day, but monitoring teams can analyze only 500 a day," says Stefan Lichtenberger, a program manager at Siemens Gas and Power. "Because some alerts are less important, while some warn of possible breakdowns or regulatory actions, these teams need help reducing and prioritizing alerts.”

Projects to reduce the volume of alerts can require two full-time employees for six months, and—because the volume of alerts inevitably rises again—must be repeated every two to three years.

"With the serverless AWS platform, we decreased customer control system alerts 90 percent and reduced infrastructure costs 85 percent."

-Stefan Lichtenberger, Program Manager, Siemens Gas and Power

A Serverless Approach Cuts Through the Noise

To help Siemens customers address this problem cost-effectively, Lichtenberger's team wanted to provide automated alert monitoring, analysis, and actionable alert-reduction recommendations. The team worked with AWS Professional Services consultants on a solution that would be called the Siemens I&C Monitors and Advisors.

"Just one month after our first meeting with the AWS consultants, we had a prototype of a serverless platform for this solution," says Lichtenberger. "Within three months, we had a prototype of the application.”

I&C Monitors and Advisors uses AWS Lambda for analyzing power plant process-control data.

The serverless solution also uses Amazon Simple Storage Service (Amazon S3) for object storage, Amazon Simple Queue Service (Amazon SQS) for fully managed message queueing, and Amazon DynamoDB to hold the results of analytics processes run against data held in Amazon S3.

"With the serverless AWS platform, we decreased customer control system alerts 90 percent and reduced infrastructure costs 85 percent," says Lichtenberger. "In the 18 months since we've been serverless, we haven't had one minute of unplanned downtime."


Learn More

Learn more about AWS serverless computing.