Zeitview

Meet Our Client

Zeitview captures inspection data about renewable energy infrastructure to help power plants and properties in 60 countries understand how and where to improve their operations and optimize manpower.

By using drones, aircraft, and an in-house analysis team to capture visual inspection data about solar panels and wind turbines, Zeitview gives clients the information they need to improve quality, mitigate risk, and boost yield.

Their Challenge

Zeitview’s data analysis team examines inspection images from all data sources, searching for visual signals about:

• Asset performance
• Corrective maintenance requirements
• System damage
• Changes in renewable energy conditions
• Progress on new system construction

This information is shared with the customer so they can decide where to focus to improve efficiency and reduce risk.

Because commercial assets like solar panels and wind turbines are so valuable, Zeitview wanted to reduce data-analysis mistakes that could cost clients $2+ million, such as the oversight of a crack in a turbine blade.

To do this, it integrated machine learning (ML) and artificial intelligence (AI) into the process. Although Zeitview has an in-house data analysis team, those workers produce customer-facing information—not data to help AI develop intelligence.

“We needed labeling services that would help us quickly and accurately label inspection images for machine learning so AI could help our team confidently identify anomalies faster,” says Jonathan Lwowski, lead AI/ML engineer at Zeitview.

The models also needed to launch within a specific timeframe. As Zeitview prepared for the North American Solar Scan (NASS), which involved inspecting every U.S. solar site that’s 1 MW or larger, the company needed accurate AI support.

Our Solution

As Zeitview rapidly scaled, they needed support to label images for ML, requiring a third party to take on this burden.

“That’s why we reached out to CloudFactory,” explains Lwowski. “They knew how to label for machine learning. They had an entire team that could help.”

CloudFactory integrates a trained workforce with a cutting-edge data annotation platform. Highly trained and qualified analysts labeled images for ML, and this information was fed to Zeitview so it could be used to train models.

As the data rolled out into production, Zeitview evaluated AI’s performance and reported back to CloudFactory so improvements could be made.

“We didn’t have to constantly QA the results or update our instructions,” says Lwowski. “They took a hands-on approach and did most of the work for us.”

The Results

Thanks to CloudFactory, Zeitview completed the NASS on time and is delivering inspection data to customers faster than ever before, achieving time-to-market improvements of six months or more and helping clients avoid revenue losses of $2+ million due to overlooked asset damage.

Zeitview successfully incorporated artificial intelligence and ML to ingest and analyze data and provide insights to customers that they can act on immediately.

“The models we have in production today would’ve taken us much longer to build on our own and would’ve required more upfront costs,” says Lwowski. “I don’t think there was a faster way to get the data labeled with the quality we needed—and within the time we needed.”