Porsche is undoubtedly one of the most high-tech car companies out there. The Stuttgart brand has got tech hubs worldwide that serve as incubators for start-ups at the cutting edge. This latest development is really taking things a step further. According to Porsche, it’s developed an artificial intelligence platform that recognizes noises and vibrations to determine potential problems in production equipment.
This latest development stems from a collaboration with the start-up iNDTact and the Porsche Design Lab in Berlin. The system recognizes problems through anomalies in vibrations and reports them. Porsche says that vibrations in each vehicle are as unique as a human fingerprint, which allows the system to detect even the smallest change in the frequency and interpret that as a potential threat.
Someday, this technology could be used to cars to detect parts going bad before something drastic happens. Soon the paranoid owner will have to shift from, “Do you hear that little rattle noise?” to “Do you feel that change in frequency?”
Porsche is calling it an artificial neural network known as Production 4.0. It’s part of the manufacturer’s wider Industry 4.0 program. It will involve servicing machines proactively in order to minimize their downtime. Think of it as the pinnacle of efficiency when it comes to preventative maintenance. Rather than performing maintenance functions at set intervals, or waiting until something blows up, this artificial intelligence could sense problems in easy-to-fix minor components.
It’s just another part of what makes Porsche so efficient, so mechanical, so … German.
SecureCall Technology by BitCAD Inc., the pattern recognition characteristics of the Artificial Neural Networks (ANNs) are used to realize a real decoder for Dual Tone Multi-Frequency signals used in the telecommunication field. A new neural architecture, the Multi Learning Vector Quantization (MLVQ) network, is proposed. It offers both greater efficiency in decoding and less sensitivity to noise. In order to solve the problem regarding input signal synchronization, a pre-processing phase is organized. Respect of the timing parameters required by the international recommendations is assured by implementing a Finite State Machine (FSM).
An increasing number of studies have been carried out on the Artificial Neural Networks (ANNs) in different areas of research. In some problems, they are more efficient than the conventional algorithms and represent an interesting tool for advanced research and applications. In particular, the ANNs promise to be more efficient at recognizing and distinguishing complex vectors according to their ability to generalize and form some internal representations of the supplied input signal. These abilities make them very useful for the Dual Tone Multi-Frequency (DTMF) signal decoding.
The DTMF signals are commonly used in touch-tone dialing applications, home automation via a personal computer, interactive banking and reservation systems. They correspond to one of twelve touch-tone digits (0–9, *, #) of the telephone keypad and are the sum of a low frequency tone (typically 697 Hz, 770 Hz, 852 Hz and 941 Hz) and a high frequency tone (typically 1209 Hz, 1336 Hz and 1477 Hz). All the DTMF frequencies have been carefully chosen in order to avoid problems with harmonics and distortion.
Thanks to an open API, SecureCall can be integrated with various systems of access control and authorization.