I. Algorithm Design

The first step in the design process is to understand what the sound analysis required should be. The first thing which we would ask any potential customer for is either pre-recorder audio samples or a permission for us to collect them. Then we will use our models in order to modify the existing algorithms or to design a new algorithm to process this sound. When this step is done, we can create an embedded system that will implement these algorithms. 

II. Data Acquisition Equipment

The other part of our work is designing data acquisition equipment. In some cases, a built-in microphone may be enough, but more often a special care needs to be taken in order to collect good sound samples.

III. Embedded System Design

When we have a reference model, embedded system design can be performed. This include designing sensor interface in FPGAs, data pre-processing and accelerator design in hardware and programming the processor for the specific application. 

IIII. Integration, Deployment and Support

When the sensor is ready, it needs to be connected to the external world. There are two ways how the sensor may be used:

  1. In a localized system, the sensor can control the equipment by setting values on its pins, which can then be read and acted on by a microcontroller.
  2. A sensor can be connected to a server either by ethernet or wi-fi connection and send the measurement information to a server.

Both approaches can be used at the same time, and the integration software is provided by ZoTech as well.