The information we present is consistently reliable. Our clients trust SAM4 to accurately determine the health of equipment and can use the information it provides to make informed decisions.
Our solutions are easy to install and maintain. They operate reliably across a wide variety of motor and asset deployments under harsh and hospitable conditions.
Our solutions are easy to install and maintain - and can be deployed across a large variety of motors and assets under both harsh and hospitable conditions.
SAM4 is intuitive and easy to use. With SAM4 you get the insight you need to prevent downtime by fixing problems before they occur.
We started with a simple idea: if we applied algorithms to the ever-growing trove of machine data, we could detect machine failures and allow our clients to prevent unplanned downtime.
We were wrong.
Well, not in all aspects. Yes, machines do generate an ever-increasing amount of data. But the simple fact that a lot of data exists does not necessarily mean that tons of value lies in it, at least not for condition monitoring. That’s why we set out to build our own sensor-system, which would create high-quality data specifically for condition monitoring. A quick online search to determine the type of data best-suited for that purpose taught us that vibration is the way to go.
We were wrong…again.
But again, we learned something valuable. Installing vibration sensors on rotating equipment in an industrial environment is cumbersome. What’s more, our sensors often failed due to harsh conditions and daily cleaning activities. In short, while vibration sensors provide high-quality data, they lack robustness and the ability to scale in our clients' demanding environments.
Inspired by our work on railway switches, which we monitored using electrical signals, we decided to work with current sensors. After encouraging test results, we decided to build a bespoke datalogger to keep the costs at an acceptable level. Prototype 1 and 2 performed remarkably well in our field lab. We set out to install 250 sensors dubbed "GEN3" at our clients' premises and expected the same results.
For a third time, it didn’t quite work.
GEN3 did manage to detect upcoming failures at an early stage. However, it generated false positives too. And many of them. Electrical waveforms contain a lot of information about the condition of motors and assets because changes in vibration or electrical failures introduce changes in patterns of those waveforms. Mechanical failures cause changes in vibration, but changes in load also cause them. So, we needed to be able to determine a load of a motor to eliminate the false positives.
Back to the drawing board. We dubbed the resulting datalogger SAM4 and included voltage probes. Having learned from earlier mistakes, we tested SAM4 in the lab and extensively in the field before deploying in live environments. In April 2018, SAM4 was ready for implementation at our clients' sites. We expected good results.
And finally, we were right.
SAM4 is performing amazingly and has detected mechanical and electrical failures up to months in advance. It has proven to be a reliable condition monitoring solution across a variety of industries. It monitors motors, pumps, compressors, conveyors, escalators, fans, and hammer mills. SAM4 provides peace of mind for asset owners, maintenance service providers and original equipment manufacturers.
The bottom line: it just works.
Founder & CCO
Head of Data Science
Head of Development
Head of Operations
University of California, Berkeley
Tilburg University, Tilburg
University of Twente, Enschede
Copyright © 2020 Semiotic Labs BV. All rights reserved.