Machine Learning and sound processing at Porsche: Hi, I’m Karthick!
Karthick Perumal introduces his job as a Data Scientist and explains in what way he makes machines learn as one of the Artificial Intelligence Specialists at Porsche.
At Porsche Digital Lab, a vital part of innovation is found in the act of learning. This is not only true for our employees, but for our technology as well. This act of machine learning is also where Karthick Perumal comes into play: As one of our Artificial Intelligence specialists, his expertise lies in creating systems that learn and improve over time and as data is added.
When he’s not creating innovative use cases and solutions during Blockchain- or AI-themed hackathons, Karthick specializes in sound-based analytics and building chatbots based on natural language processing. The applications of these fields are vast, even when limited to the domain of Porsche.
They range from improvements to the sensors for autonomous cars all the way to predictive maintenance systems based on acoustic fingerprints. One such case is the Sound Detective, where an AI was exposed to vast amounts of acoustic data generated by a machine through its daily use.
Through machine learning, the AI was soon able to identify the sounds generated when all the components were working perfectly, and whether any component was worn down or malfunctioning. This AI can look for any anomalies in the sound and is able to tell, based on sound input alone, when the machinery — in this case, a coffee machine — might breakdown and what parts need to be replaced or fixed soon in order to avoid any potential downtimes. Thus, reducing operating costs.