AI-driven platform predicts mechanical machine errors

Augury’s predictive maintenance platform uses artificial intelligence to analyse machines for mechanical errors

Spotted: The self-described “machine health” startup Augury has developed a predictive maintenance platform that uses artificial intelligence to analyse machines for mechanical errors. The idea is that readings and patterns embedded within the noise from motors, compressors, pumps, industrial-scale heaters, etc., can be used to detect a problem. 

Augury’s sensors record the readings and process the vibrations, temperatures and magnetism metrics of the machines, before uploading them to the cloud to be analysed by AI algorithms, which are generated by baseline readings in the cloud backend. The system gradually begins to recognise abnormal sounds and faulty movements and the machines analysed are then compared to similar appliances on the cloud, relieving the need to retrain models.

The technology can be scaled up, and the company has expanded from analysing pumps, fans and chillers, to noncritical machines. The startup recently secured $55 million in a funding round, which they say will be used to “sustain and accelerate Augury’s growth through new hires, R&D efforts, and customer acquisition in the US and beyond.” Indeed, the global pandemic seems to have aided, rather than haltered their expansion.

Explore more: Computing & Tech Innovations | Work & Lifestyle Innovations