Iot Analytics

Industrial IoT Analytics Tech Is Moving To ‘The Edge’

Pinterest LinkedIn Tumblr

Industrial IoT Tech is moving to the edge.

Low-code Application Development Company

I suppose it makes sense that cutting edge industrial ‘internet of things’ technology is increasingly happening on the edges – meaning on the far edges of networks, often directly on devices.

Rather than moving mountains of data directly from devices into the cloud, where it can be collected, analyzed, and – presumably – put to good use, more and more grunt work of IoT is being handled by software-enabled hardware attached to equipment.

That’s one takeaway at LiveWorx, PTC’s IoT conference this week in Boston.

“Over 40 percent of (IoT) processing will eventually be done on the edge,” Deloitte Chief IoT Technologist Robert Schmid told LiveWorx attendees. “There’s going to be processing going on at the edge, and it’s going to be great.”

While this may seem inefficient on its face – isn’t it simpler to ship everything to the cloud where it can be analyzed and acted on by a central platform? – it actually solves two important problems.

First, industrial IoT users are very worried about security. They don’t want data about their output, processes and technology to get into the hands of their competitors. And they definitely don’t want malicious hackers disrupting their operations, or using IoT networks to find a backdoor into financial data. By putting hardware “at the edge,” users create a potentially more secure barrier between their devices and data, and the outside world.

Second, edge computing makes analytics easier by filtering data, more often allowing it to be understood and acted on in manageable amounts, directly on the device.

Andrew Timm, VP of Technology Platform at PTC, offered a striking example of this benefit for Tuesday’s LiveWorx crowd. One typical device on one typical factory floor can generate up to 50,000 sensor readings per second, he said. Expand this to an entire factory floor, then to a network of factory floors, and a large industrial manufacturer can quickly find itself processing more than 5 billion sensor readings per second, he said.

Timm, demonstrating PTC’s ThingWorx platform, explained that a device on the edge can analyze and learn from data on it’s own, building its knowledge of the device.

“It quickly learns what normal is,” Timm said.

This means it learns what abnormal is, of course, allowing it to take action and alert operators. It also means that less information needs to be sent into the cloud, where the user’s network resides.

“You can go from 50,000 (sensor readings) down to 100 alerts per day,” he said.

ThinkDataAnalytics is a data science and analytics online portal that provides the latest news and content on AI, Analytics, Big Data, Data Mining, Data Science, and Machine Learning. A team of experts with extensive experience in the field runs