Transforming Vast and Underused IoT Data into Insights for NIBE Group

In Brief

• NIBE Group faced challenges in managing and leveraging its vast and underused IoT data.
• Redeploy implemented a comprehensive data management solution using Databricks, Spark, and Delta Lake.
• The result was accelerated analytics and improved operational effectiveness.

Their Challenge

NIBE Group was dealing with a vast amount of IoT-generated events data stored in many non-queryable small files, making it difficult to identify and understand IoT device alarms. For example, they wanted to determine optimal filter replacement schedules based on airflow.  

Ultimately, NIBE Group sought to gain deeper insights into their heat pumps and eventually implement a data platform to enable the capabilities.

“To stay at the forefront of our industry, we had no choice but to find a way to handle our massive data storage blob more efficiently. Through a recommendation from Microsoft, we contacted Redeploy, who implemented a whole new approach to data and the capabilities to predict maintenance and generate more value for our customers.”

– Kenneth Linden Magnusson, Chief Sustainability Officer at NIBE Group

Our Solution

Our Redeploy Insight team implemented a comprehensive solution leveraging several key technologies:

• Databricks and Spark were used for ingestion by utilizing massive and parallel compute capabilities, allowing NIBE Group to handle the vast volume of IoT data effectively.
• ‍Delta Lake provided a scalable storage solution to turn billions of small files into an efficient format.‍
• Databricks Autoloader was implemented for efficient data loading – streamlining the process of ingesting IoT data.
• Data correction techniques, including Large Language Models (LLM), were employed for data correction and standardizing categorizations of data fields, thereby enhancing data accuracy and completeness.

The solution provided data management of heat pumps across various areas, types, and classifications – identifying and analyzing common issues. The structured sensor data has been utilized for Business Intelligence (BI) reporting, machine learning POCs, and monitoring of various IoT sensor parameters within pumps.

The Results

NIBE Group’s enhanced data management capabilities allow them to proactively address device support measures and accelerate analytics to improve next-generation devices and firmware updates for existing pumps, refine maintenance strategies, and educate operators.      

This comprehensive approach to addressing device issues results in quicker and more effective client support, ultimately delivering greater value to them.