Hardware all along

We talk about Machine Learning, Artificial Intelligence, Data Science, Data Analytics as some of the leading technologies in the market at the moment. While these can be applied to various domains, for example, Manufacturing or Automotive, the ultimate low level data for all these has to be generated from some point and that point is obviously the hardware. When we say hardware, it can be a simple sensor read and transmitted to a cloud using device like Gateway. Interfacing a sensor or any other device needs hardware design. At the core level to generate the data that is to be understood by Gateway, the data has to be formatted which is done by the chips like micro controller. The micro controller has to be programmed and data to be transmitted to Gateway over communication interfaces like UART, Ethernet, etc. While these are wired interfaces, these can be extended to wireless interfaces using transceivers. The Wireless interfaces could be Bluetooth, Wi-Fi, Zigbee, Short Range Wireless communication. So, until the point data is provided to the ML algorithms it is all the hardware who has to do the job. While the role of ML/AI could be writing complex algorithms, they must have basic idea of what is happening at the hardware level.