Back to search

BIA-Brukerstyrt innovasjonsarena

On-Edge Anomaly Detection in Machinery Using Sound as a Data Source

Alternative title: EARONEDGE

Awarded: NOK 16.0 mill.

Project Manager:

Project Number:

337085

Project Period:

2022 - 2025

Funding received from:

Maintenance is the most resource-intensive part of managing a large building. Today, a large part of maintenance is done “the old fashioned way” by pre-planning a maintenance schedule and sending technicians for regular check ups. The result of this is that a lot of valuable time is spent inspecting rooms and machinery that don’t need it and irregularities or machine faults often come as a surprise to both tenants and property owners. A major part of a technician's inspection is to listen for sounds that are out of the ordinary (anomalies), such as whining or thumping sounds from a pump This project is an effort to develop and train algorithms to mimic human hearing, which, when placed around in various parts of the building, will act as an omnipresent technician that can detect any sign of irregularities

In this project we will develop technology that will use sound detection to provide real-time information about machinery status to property managers, automatically inform about irregularities in their equipment without manual “checks”, and allow for more effective maintenance planning. Maintenance is the most resource-intensive part of managing a large building. Today, a large part of maintenance is done “the old fashioned way” by pre-planning a maintenance schedule and sending technicians for regular check ups. The result of this is that a lot of valuable time is spent inspecting rooms and machinery that don’t need it and irregularities or machine faults often come as a surprise to both tenants and property owners. A major part of a technician's inspection is to listen for irregularities (i.e., anomalies). This project is an effort to develop and train algorithms to mimic human hearing, which, when placed around in various parts of the building, will act as an omnipresent technician that can detect any sign of irregularity. The exploitation potential of the innovation described in this project can be seen in the context of three particular current market trends: 1. The expectation of continued explosive growth in the IoT market over the next 5 years 2. The recent introduction of advanced AI-optimised chipsets enables a whole new dimension for edge-optimised processing 3. Automation and digitisation of infrastructure management has already become a core part of high-yield, capital-intensive operations like oil rigs and advanced processing facilities, and is now increasingly democratised and additionally made available for lower-cost applications (like property management)

Funding scheme:

BIA-Brukerstyrt innovasjonsarena