Description
In the age of mobile cloud computing web-based systems are often designed to transfer data to large scaling online storage facilities in order to persistently save and analyze it with complex algorithms such as used in machine learning. These systems often require a reliable network connection, which does not hold for a variety of mobile business applications. As an alternative to traditional cloud-based systems the TUCANA approach makes use of the local processing power of mobile edge devices in order to come up with high complex AI pipelines processing data in real-time. By applying the idea of TUCANA to our service use-case called “nPotato” we developed an artificial, nociceptive potato that frequently measures and analyses acceleration data during the harvesting process of potatoes. In the given scenario sensory data is processed locally in real-time using the device’s local computing power to gain higher productivity in the area of precision farming.
TUCANA: A platform for using local processing power of edge devices for building data-driven services
In the age of mobile cloud computing web-based systems are often designed to transfer data to large scaling online storage facilities in order to persistently save and analyze it with complex algorithms such as used in machine learning. These systems often require a reliable network connection, which does not hold for a variety of mobile business applications. As an alternative to traditional cloud-based systems the TUCANA approach makes use of the local processing power of mobile edge devices in order to come up with high complex AI pipelines processing data in real-time. By applying the idea of TUCANA to our service use-case called “nPotato” we developed an artificial, nociceptive potato that frequently measures and analyses acceleration data during the harvesting process of potatoes. In the given scenario sensory data is processed locally in real-time using the device’s local computing power to gain higher productivity in the area of precision farming.