Modern IoT data pipeline

Double exposure of businessman shows modern technology as concept

With the emergence of cloud computing technologies, IoT solutions are increasingly easy to deploy at scale and companies can reap considerable benefits with relatively small investment. IoT enables businesses to have real-time data from all the levels of their value chain, enabling multitube of things, such as more accurate decision-making, better customer orientation, agile reaction to anomalies, etc.

IoT solutions are a natural fit for BDS Bynfo, whose mission is to help companies in implementing modern cloud-based solutions with a technology stack that contains most proven technologies. Next paragraph provides a photogenic example of a modern IoT solution that BDS Bynfo has set up.

The case company is a large manufacturer with capital-intensive production machinery running around the clock. Avoiding machine downtime is paramount for the company, as most of the costs occur from these machines. Before, the machines would be routinely offline from time to time, which results in large losses for the company, in form of lower utilization rate.

Solution was found from IoT with the help of BDS Bynfo. Coupled with hundreds of sensors, the machines send thousands of telemetries per minute containing information about their state and production conditions. This data is automatically pipelined to go through cloud all the way to the end-user dashboards. Some of the features and benefits of this setup include:

  • ML model that predicts machine downtime hours ahead with impressive accuracy;
  • Real-time monitoring of the machinery;
  • Automatic and flexible pipeline that adjusts its requirement for computation power to match the amount of incoming data;
  • Cost-efficient way to store and visualize the data--> Savings, predictability and better decision-making!

This real-life example is put to a demo, where we simulate a single smart sensor of an imaginary network operator, that wants to have real-time data from its customers and react pre-emptively to anomalies. The point is to demonstrate the ease of setting up an easily scalable IoT solution, with the help of modern cloud technologies.

The demo architecture is deployed to Azure with Terraform. We are using Raspberry Pi to simulate the sensor. Snowflake is used for the central analytical database to store, transform and share all the data in this demo. This setup works equally well with other cloud platforms and scales easily to match the real-life needs, where the amount of data and sensors would be significantly higher.