In this episode Andrew and Danny look into why measuring the temperature in supermarket fridges is really important, and it turns out to be amazingly interesting!
Synopsis:
In this podcast we talk about an example where a supermarket chain can automate the collection of temperature readings in their fridges. Yes, really! It’s not as dull as it sounds!
Hundreds of branches of a supermarket, with tens or hundreds of fridges in each store, measuring temperature every second or so makes a whopping amount of data.
We view this as a commercial IoT application, and build solutions for this using tools such as Azure IoT Hub.
Dealing with the data is typically done in two ways:
Firstly, we store all the data in a large data store such as a data lake for analysis later. When we want to analyse historical trends, or get statistics over an extended period of time, this is what we do. We’ll use a tool such as Databricks – built on Apache Spark – to crunch big datasets and then use Power BI or Tableau to report and dashboard the results.
Realtime analytics would be handled differently. We’d be putting the data messages onto a stream such as Apache Kafka or Azure Service Bus and then feeding this into a streaming analytics solution. This could again be something such as Databricks.
Rounding off, we talk about the amazing power of Cloud. In days gone by the level of investment to put a big data solution would be huge and take many months. Now we can spin up a data lake and Databricks in minutes and get a solution together in hours. It’s incredible.
Stay tuned for more fun next time!
Watch the video here:
Listen to the audio here: