example described: "smart traffic intersection. ... four video cameras connected to an Internet of things (IoT) device with a small CPU, similar to a Raspberry Pi. The cameras send video frames to the IoT device, where they’re analyzed using a machine learning (ML) image-recognition model and control instructions are then sent to the traffic signals. One of the small IoT devices is connected to Azure Cloud Services, where information is logged and analyzed offline.
The demo programs described in this article (a Python program that uses the CNTK library to create an ML model, a C program that simulates IoT code and a Python program that uses an ELL model"
The controller device described in article could be Azure Sphere
Microsoft CNTK has some advantages, but is a bit less "popular" than Google's TensorFlow library.
And for Microsoft developers, having C# (Core) option may be preferred to Python.
Anyway, an interesting article.