We live in a world gone mad with sensors that now gush data, flooding all human ability to manage it or make sense of it. Most of that data runs off, ending up in vast “data lakes,” a nice name for what is really server farms in data centers, the modern-day equivalent of immense warehouse stacked with dust-collecting boxes.
However, a data stream could be very useful in detecting irregularities in a machine’s operation. With AI and its ability to learn, on its own, about what data constitutes regular operation and what might be an anomaly, an irregular signal of trouble ahead, preventative maintenance can be performed to save the day.
Like a doctor watching a cardiogram and able to spot an underlying heart condition, so could a software application. Add to that some machine learning, a library of tagged malfunctions and their respective signals, and we have an automatic, full-time sentry with lightning-fast reflexes, able to ring alarms and summon help.
Such is the promise of Falkonry, a Cupertino-based startup, recently acquired by IFS, the billion dollar service software company. Never heard of them? You will. The Swedish company has big plans for entering the U.S. market.
Read more about Falkonry and the acquisition on engineering.com.