Posted in | News | Machine-Vision

New AI-Based Method for Different Mechanical Failures Detection in a Noisy Environment

Every year, around 1 trillion dollars is being lost by the world’s largest manufacturers to machine failure. Several issues come under the noisy factory surrounding — working equipment and processes generating can be loud. As a result, machinery faults are frequently inaudible or for that reason detected too late.

New AI-Based Method for Different Mechanical Failures Detection in a Noisy Environment.
Rytis Maskeliūnas, Kaunas University of Technology, Faculty of Informatics researcher. Image Credit: Kaunas University of Technology.

Scientists from the Kaunas University of Technology (KTU), Lithuania suggested an artificial intelligence-based technique for various mechanical failures detection in noisy surroundings. The new solution seems to not only be sustainable — equipment can be easily digitalized without remodeling it — but also affordable.

In industrial machines, anomaly detection is a technique that depends on different data — pressure, temperature, vibration, sound and electric current — all from sensors installed inside the machine itself. Although sensors seem to be necessary for capturing fundamental diagnostics, they are hard to establish in older generations of factory lines as the machinery is highly “mechanical” and “not digital”.

For factories with low automatization levels, many of which remain much larger than autonomous manufacturing lines, failure detection without employing new sensors for each industrial machine is extremely important. As the sound data is easy to collect because of the relatively low installation cost of contactless microphones to existing facilities, sound data-based methods are of great interest.

Rytis Maskeliūnas, Study Co-Author and Researcher, Kaunas University of Technology

But in highly noisy factory surroundings, the sound data becomes interrupted and contaminated, leading to misinterpretation of the sounds and falsely denoted mechanical failures.

The deep machine learning (ML) method was recommended by the multimedia and software engineers’ team from KTU. It depends on real-life sound data obtained from working industrial machines and can be utilized for machine diagnostics with no needless installations of new sensors. Maskeliūnas feels that failure detection depends on training algorithms with real-life sound data inside real industrial machinery sound data.

The new software solution is cheap and easy to use—the only equipment needed is a microphone pool and a processing device. Artificial intelligence allows acoustic anomaly detection with no additional sensors.

Rytis Maskeliūnas, Study Co-Author and Researcher, Kaunas University of Technology

A Sustainable Solution to help Digitize the Industry

The purpose was to improve the robustness of anomaly detection in the domain of mechanical motion. This is a perspective field, because of sustainability and the opportunity to digitize the industry without getting rid of old equipment as new factory installations require a lot of resources and will not happen any time soon in a lot of poorer countries.

Rytis Maskeliūnas, Study Co-Author and Researcher, Kaunas University of Technology

The experiments were performed on the Industrial Machine Inspection and Inspection Malfunction Investigation and Inspection (MIMII)  a sound dataset of industrial machine sounds. Maskeliūnas feels that this data set consists of four various kinds of machinery: valves, fans, pumps and slide rails. The waveform audio file (.wav) format was utilized to store the data that consisted of machine sound and noise.

The noise is real manufacturing environment sound that was intentionally blended with pure machine sound at three different SNR — signal-to-noise — levels: 6 dB, 0 dB, and 6 dB. The machine sound was recorded for both normal and abnormal conditions. As a result, we proposed an anomaly detection system for the analysis of real-life industrial machinery failure sounds,” stated Maskeliūnas.

Machine Failures are Time-Dependent

According to Maskeliūnas, the inclusion of acoustic new sensor technologies integrated with deep learning methods can be utilized to avoid unwanted replacement of equipment, decrease maintenance charges, enhance work safety, increase the handiness of equipment and hold acceptable levels of performance.

Maskeliūnas stated, “Early warning can be obtained through the predictive maintenance system based on acoustic failures recognition. The ability to detect weak signals may have a strong strategic impact. Their key benefit is real-time management and planning, which helps to cut down on the costs of production downtime.”

The KTU scientist group plan to detect more types of failures.

 “Like most artificial intelligence researchers, we are limited by the amount of data we have. A partnership with a manufacturing company would allow us to gather different scenarios and apply the method more widely. Our solution is particularly relevant in countries with little digitization where companies do not have resources for new equipment,” remarked the team.

The novel method towards acoustic anomaly detection has received previous inquiries for implementation in industrial surroundings. Maskeliūnas observes that the biggest advantage is that it is affordable and no installation is needed — only a recording is required.

Journal Reference:

Tagawa, Y., et al. (2021) Acoustic Anomaly Detection of Mechanical Failures in Noisy Real-Life Factory Environments. Electronics.


Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.