Big-data company Cloudera Inc. today announced its first new product since the completion of its merger with former rival Hortonworks Inc., targeting workloads at the so-called network edge.

Edge data generally refers to information that’s created by “internet of things” devices and sensors. The idea behind Cloudera’s new edge data management tools is to help companies absorb and process that data at its source, then analyze it to provide immediate insights that can help to make better business decisions.

Processing data at the edge means those insights can be accessed faster than if the information is first transferred to a remote data center, processed and then returned to its source, as is usually the case.

Being able to do so is important for several reasons, Vikram Makhija, general manager of Data in Motion at Cloudera, told SiliconANGLE. First, IoT sensors and devices have traditionally had to send all of their data over a network to a remote server for analysis, but this can put an unnecessary strain on those servers and impact an organization’s ability to analyze that information in real-time.

“Having the right data at the right time is what allows for key insights to be generated in real-time,” Makhija said. “By enabling the edge to be more intelligent, by embedding and executing machine learning models at the edge, one can expect only quality and relevant data to be pushed to the enterprise for analysis. The industry is shifting towards offering analytics at the edge, and avoiding costly round trips is becoming more critical.”

Edge processing capabilities also help with the “disconnected edge,” which relates to systems located outside of a functioning network. They include things such as aircraft and trucks moving through areas where there’s no network available, or systems in industries in which regulations prevent data being transferred to remote data centers immediately.

“In all these scenarios, actions may need to be taken at the edge even though the data may not be fully available at the enterprise,” Makhija said. “For this, the edge needs to be smart and capable of understanding and processing the data.”

There are also big cost savings to be had by processing data at its source.

“It’s not just about time, it’s also about cost,” Makhija explained. “Transferring data from the edge to the data center can be costly and sending lots of raw data is inefficient. Processing at the edge allows for applications that are not just more responsive but also more efficient.”

The company’s new tools include Cloudera Edge Management, which is designed to improve the way in which IoT data is processed. Companies can set up software “agents” that collect data from remote sensors and devices and process it directly at the edge. Cloudera Edge Management also provides a “code-free” drag-and-drop environment that’s used to deploy machine learning models easily to those agents, so decisions can be made at the edge without data needing to make round trips to and from the enterprise.

The second new tool is Cloudera Flow Management, which is used to control the flow of data throughout an organization. The tool is said to reduce data integration times by enabling petabytes of data to be securely moved around data centers, cloud and on-premises environments, wherever it needs to go.

The new tools will be added to Cloudera’s DataFlow platform, which is the company’s main streaming analytics platform based on open-source software such as Apache Hadoop and Apache Spark.

“With DataFlow, Cloudera already had a differentiated offering for processing and analyzing data in motion,” said Matt Aslett, research vice president of data, AI and analytics at 451 Research Inc. “Cloudera Edge Management adds the ability to develop, deploy and monitor data processing applications at the edge, which is likely to be a fundamental enabler of successful IoT projects going forward.”

The new tools will be made generally available before the end of the March.

Image: 200Degrees/Pixabay