Google has demonstrated new Internet of Things (IoT) hardware and software that the company is hoping will transform the landscape for these products and help it better compete with the likes of Amazon and Microsoft.
Among the innovations demonstrated at Google’s Cloud Next conference was an extension of its Cloud IoT software platform to better support edge computing, Network World reports.
Edge computing refers to a network architecture where analysis and data processing operations are performed at or near to the endpoint where the data is initially gathered from, rather than being sent back to a central server to be worked on. It is considered particularly important to many IoT use cases where speed is a key factor, because it can greatly reduce latency as data does not have to be transmitted through the network.
Another advantage of edge computing, as noted by Christian Renaud, IoT research director for 451 Research, is that it can allow companies to still enjoy the benefits of IoT sensors, even if they aren’t comfortable with sending sensitive data to the cloud for processing.
Google’s move to extend its IoT software to edge devices could therefore make its offerings much more attractive to buyers and brings it in line with Microsoft and Amazon, who already offer edge computing services.
Elsewhere, Google also debuted a new chip at the Cloud Next event called the Edge TPU, which it claims uses the company’s TensorFlow AI software to improve machine learning.
This custom chip, which is comfortably smaller than a penny, has been designed specifically to run advanced machine learning models directly on endpoint devices, enabling sensors themselves to generate useful predictions and insight.
Google states that potential use cases for the chip include predictive maintenance, anomaly detection, machine vision, robotics and voice recognition.
Mr Renaud commented: “Depending on the effectiveness of Edge TPU, this may have not only caught them up, but allowed them to pull ahead of the competition technically.”