Energy asset discovery and configuration tool

The configuration of energy management solutions requires large amounts of manual data entry and relies heavily on subject matter experts. This particularly the case today because modern buildings can provide large amounts of data that is often poorly structured and labelled as they are commissioned by different people without any specific labelling standard.

Identification of relevant assets and their data points for energy management therefore ofen requires a large manual. This increases not only the installation costs of energy management systems. The additional integration effort also keeps potential customers with legacy systems from installing energy management systems.

The Building Energy Asset Discovery (BEAD) tool semi-automates the labelling process and allows smart building analytic problems such as energy management and diagnosis to be solved automatically. The approach uses semantic and artificial intelligence techniques to identify and label a building’s energy assets automatically and directly from plain sensor lists.

By extension, semantic reasoning is used to derive a detailed model of the physical cause-and-effect relationships between sensors that can be used to configure analytics and diagnose anomalies in buildings.


Joern Ploennigs


Deploying analytic and management solutions in buildings is challenging due to


Example of result

For example the Dublin Technology Campus with 6 buildings and 9,820 sensors was analyzed. The BEAD tool was able to label the dataset within 5 minutes and create a configuration for the IBM energy management software.

The extracted cause-and-effect relationships between sensors led to a higher coverage of detectable and diagnosable behavior. The diagnosis model computed for the campus covers 2,411 sensors with 1,446 diagnosable anomalies resulting from 47,284 potential causes. It allowed several anomalies in the building operation to be detected and diagnosed.