Preventing AHU Belt Failure - Shoreline | Industrial Automation

Timely alerts from Shoreline APM prevents belt failure of the Air Handling Unit

Early Detection of Belt Misalignment Averts Major Equipment Damage

Total vibration energy (TVE) on Motor DE bearings rose from 0.45 ips to 0.85 ips, which are considered high for a motor/powertrain this size. The vibration spectrum showed high 0.88x Motor rpm illustrating high vibrations from fan (fan rpm = 0.88x Motor rpm) were showing up at Motor DE bearing.
Alarm: A standard-deviation alarm was triggered based on the increase in total vibration energy (TVE) and high 0.88x Motor rpm vibrations.
The presence of strong 2nd and 3rd harmonics in the vibration spectrum pointed towards belt misalignment and/or correction of belt tension.

Real-Time Monitoring of Assets by Shoreline AI

Supporting Data

The long-term trend below showed no gradual increase, i.e. the trend indicates slow and persistent development of a machinery fault, and was persistent for several months.

Evidence of Success:

Armed with this precise insight, the maintenance team took immediate action by replacing the belts and realigning the sheaves. The results were striking. The vibration spectrum showed a dramatic reduction in the problematic frequencies, with TVE dropping from nearly 1 inch/sec to a healthy 0.52 inches/sec. The motor was back to normal operation as shown by the before and after repair trends below.

Data ingestion and processing into Shoreline’s AWS environment consists of several AWS services that include AWS IoT Core which handles device authentication/authorization, data encryption. IoT core triggers AWS Lambda for processing. Structured data is stored in Amazon Aurora while unstructured data and asset manuals in stored in Amazon S3. Shoreline also leverages Amazon SageMaker for training its ML models and Amazon Bedrock LLM Models for Gen AI. This combination of ML models with physics-based models enable auto-configuration of assets from Shoreline’s proprietary pre-built library of more than 30,000 assets. Initial machine baselines established from the auto-configuration process provides insights in days, not months.

Conclusion:

This case demonstrates the power of predictive maintenance. By quickly identifying the need for belt replacement and sheave realignment, the system’s normal operation was restored, and a costly equipment failure was avoided, saving both time and money.

About Shoreline AI

Shoreline AI’s plug-and-play asset performance management delivers breakthrough simplicity and cost efficiencies. Completely self-installed by non-experts, smart sensors automatically connect to the cloud and are auto-provisioned via a rich library of 30,000+ pre-built asset physics models.

This cloud-native approach requires no new CapEx, on-site experts or data scientists, operationalizing in days and delivering powerful machine-specific analytics. This highly secure, 100% subscription approach creates unprecedented industrial APM economics and scales easily for new applications such as emissions monitoring.

Shoreline AI helps clients in asset-intensive industries maximize the performance and profitability of their operations, create a proactive and predictive approach to asset management, and accelerate sustainability initiatives. The company’s solutions are designed for machinery serving the energy, manufacturing, pharma and data-center cooling industries.