It was a dark, damp night. Not a soul in sight. Somewhere in the distance, a clock ticked and a chime whistled in a coordinated rhapsody. Every once in a while, the harmonics of this orchestra seemed to be interrupted by a shrill ‘high-E’. It just didn’t sound right…
Right, there is a reason why I am not an acclaimed suspense author. Nor is the paragraph above meant to be a preview of my yet unpublished Great American Novel.
Industrial IoT (IIoT) systems have existed for many years. They are rock stable, and work well. They are also from an era where ‘connectivity’ was limited to none, and smartphones, alexa, push notifications et. al were non-existant. The “shrill high E” I talked about in my novel attempt might just be a sign for a warning in the IIOT space. A big warning. It might be a pressure valve giving way. It might be a pin hole liquid leak.
Some of the key differences between “consumer IoT” and “Industrial IoT” have to do with resiliency and impact of failure. I recently attended an IIOT conference where someone joked “if your Internet coffee machine doesn’t download a recipe, you get mildly annoyed and you hashtag their support on Facebook. If a gas pipe explodes because a sensor did not catch it in time, people die”. Obviously ,this was trivializing the impact of consumer IoT breaches (it can get much more serious that that), but the larger message that IIOT devices and sensors have a more material impact on failure is generally true.
This is also one of the big reasons why industrial IoT companies don’t want to add a lot of intelligence to sensors that work and work well. It involves a lot of cost to retest and make sure they adhere to stringent standards. Furthermore, every new ‘connectivity’ technology involves adding new protocol stacks, impact on per unit cost, impact on battery (if applicable) etc.
This is exactly where oZone comes in. We’ve built out technology that allows IIoT companies to connect their unconnected sensors without product re-engineering costs. In other words, we create a ‘bridge’ for them for legacy sensors.
We do this with external cameras and audio sensors. With some intelligent training, our code can “automatically read” sensors that have no connectivity support. We also have an audio framework that can be trained with audio samples on what might be a ‘normal’ operation vs ‘abnormal’ conditions. We run a variation of anomaly detection algorithms and help IIOT administrators get “remote notifications” and “monitor alarms” on legacy devices.
Talk to us, if you are an IIOT company and this looks like a cool idea.