AiProd: the artificial intelligence application of Acoustic Insights
New and exclusive predictive features in manufacturing contexts
AiProd is the PATENTED solution, proprietary solution of IProd Srl, that uses the Acoustic Insights technology of IBM to realize the integration of IoT data with audio files acquired by machines, enabled by the Alleantia I4.0 "Plug&play" technology and IBM Watson in-cloud artificial intelligence system for the recognition of the acoustic pattern.
The solution, capable of predicting results on the quality of the machining, or of a product being tested, or even machine failures and defects, opens up interesting applications that may help manufacturers and end-users as well to prevent costly production waste or downtime for technical assistance.
AiProd is the result of the collaboration of three exclusive components. In particular:
Alleantia provides both the I4.0 “Plug&Play” interconnection technology of machine tools and production systems’ controllers, and a new functionality of multimedia peripheral drivers (microphones and even cameras), which strengthen the breadth of its already powerful library of over 5,000 drivers of industrial devices ready to use.
IBM, with Watson's Acoustic Insights function, provides an Artificial Intelligence system for training, refinement and recognition of the acoustic pattern. The technology compares reference audio files, or those known to be representative of the optimal functioning of the machines, with new audio files acquired in real time, giving it a score of equivalence between 0 and 100%, where the 0% score is equivalent to a completely new sound and 100% to an identical sound already known by the system.
IProd offers the innovative application AiProd “powered by IBM” that combines, in real time, multimedia files associated with IoT tags of production assets with Artificial Intelligence patterns. The latter create clusters capable of training a necessary and sufficient number of neural networks to ultimately perform increasingly accurate surveys and be associated with homogeneous processing scenarios.
