Using AI based acoustic monitoring to detect pipeline or machinery anomalies
The Problem:
There are often long delays between an asset anomaly occurrence and its detection.
High false positives also result in increased manpower costs.
The Solution:
Teredo Analytics’ pipeline listener and machinery listener uses non-invasive methods to detect anomalies. The combination of event-triggered monitoring and trained analytics engine makes the products unique and >90% accurate.
The Differentiator:
Teredo is focused on providing a comprehensive asset monitoring solution via its three products, whether it be pumps in a room or a pipeline network.
We use directional acoustics and ultrasonic technology to provide insights about pipelines and different network assets. We make use of event-triggered algorithms to only trigger sensors when needed, which helps in providing multiple year long battery life.
We also have a trained Analytics engine that sits in our secure cloud and works as an additional decision layer to reduce high false positives. Both our co-founders come from Acoustic Research Lab of NUS and have extensive experience in deploying sensors & robots in the field. Last year alone, a single installation of our sensor helped PUB in direct cost savings worth 80k SGD and prevented an accident that could have cost them 200k SGD.
Biggest Achievement:
1. Teredo currently have 3 paid projects live with PUB. We have been proven to save 390 man days and $78k per year.
2. POC with SMRT, starting in August to monitor elevators.
3. Teredo was one of the 15 finalists that were awarded 100k by an esteemed panel of judges at NUS GRIP.
4. Teredo has received grants from PUB to develop its water infrastructure sensors.
We look forward to starting POC trials with SMRT at three MRT stations in Singapore.
